HDL Measures, Particle Heterogeneity, ProposedNomenclature, and Relation to Atherosclerotic

Cardiovascular EventsRobert S. Rosenson,1* H. Bryan Brewer, Jr.,2 M. John Chapman,3 Sergio Fazio,4 M. Mahmood Hussain,5

Anatol Kontush,3 Ronald M. Krauss,6,7 James D. Otvos,8 Alan T. Remaley,9 and Ernst J. Schaefer10

BACKGROUND: A growing body of evidence from epide-miological data, animal studies, and clinical trials sup-ports HDL as the next target to reduce residual cardio-vascular risk in statin-treated, high-risk patients. Formore than 3 decades, HDL cholesterol has been em-ployed as the principal clinical measure of HDLand cardiovascular risk associated with low HDL-cholesterol concentrations. The physicochemical andfunctional heterogeneity of HDL present importantchallenges to investigators in the cardiovascular fieldwho are seeking to identify more effective laboratoryand clinical methods to develop a measurementmethod to quantify HDL that has predictive value inassessing cardiovascular risk.

CONTENT: In this report, we critically evaluate the di-verse physical and chemical methods that have beenemployed to characterize plasma HDL. To facilitate fu-ture characterization of HDL subfractions, we proposethe development of a new nomenclature based onphysical properties for the subfractions of HDL thatincludes very large HDL particles (VL-HDL), largeHDL particles (L-HDL), medium HDL particles(M-HDL), small HDL particles (S-HDL), and very-small HDL particles (VS-HDL). This nomenclaturealso includes an entry for the pre-�-1 HDL subclassthat participates in macrophage cholesterol efflux.

SUMMARY: We anticipate that adoption of a uniformnomenclature system for HDL subfractions that inte-grates terminology from several methods will enhanceour ability not only to compare findings with differentapproaches for HDL fractionation, but also to assess

the clinical effects of different agents that modulateHDL particle structure, metabolism, and function, andin turn, cardiovascular risk prediction within theseHDL subfractions.© 2010 American Association for Clinical Chemistry

Conventionally, cardiovascular prevention strategiesemphasize therapeutic reductions in LDL cholesterol(1, 2 ). However, increasing attention is being focusedon HDL cholesterol as a secondary prevention target toaddress residual cardiovascular disease (CVD)11 risk(3, 4 ). Low HDL cholesterol is widely prevalent inWesternized countries, and is independently predictiveof CVD risk (5, 6 ), even at low concentrations ofLDL cholesterol (7 ). However, low concentrations ofHDL cholesterol are often accompanied by increasedconcentrations of small, cholesterol-depleted LDL par-ticles and increased concentrations of cholesterol-enriched triglyceride remnants. Thus, the CVD risk as-sociated with low HDL cholesterol values is difficult toseparate from that of other associated lipoprotein ab-normalities (8 ).

HDL particles are heterogeneous in size and com-position. Despite the substantial epidemiological datasuggesting a cardioprotective role for HDL, much re-mains unknown about the antiatherothrombogenicproperties of different particles that comprise this classof lipoproteins. Some of the attributes of HDL inmitigating atherosclerosis include its role in reversecholesterol transport, oxidation, and inflammation(9, 10 ). Several genetic mutations can affect the struc-

1 Mount Sinai Heart, Mount Sinai School of Medicine, New York, NY; 2 MedStarResearch Institute, Washington DC; 3 INSERM Unit 939, UPMC Paris 6, Hopitalde la Pitie, Paris, France; 4 Vanderbilt University, Nashville TN; 5 SUNY Down-state Medical Center, Brooklyn, NY; 6 Children’s Hospital Oakland ResearchInstitute, University of California, Berkeley; 7 University of California, San Fran-cisco, CA; 8 Liposcience, Raleigh NC; 9 Lipoprotein Metabolism Section, Pulmo-nary and Vascular Medicine Branch, National Heart, Lung and Blood Institute,National Institutes of Health, Bethesda, MD; 10 Lipid Metabolism Laboratory,Tufts University, Boston, MA.

* Address correspondence to this author at: Mount Sinai School of Medicine, OneGustave L. Levy Place, Box 1030; New York, NY. Fax 212-659-9033; e-mailrobert.rosenson@mssm.edu.

Received September 16, 2010; accepted December 3, 2010.Previously published online at DOI: 10.1373/clinchem.2010.15533311 Nonstandard abbreviations: CVD, cardiovascular disease; CHD, coronary heart

disease; 2D, 2-dimensional; VAP, vertical auto profile; apo AI, apolipoproteinA-I; ABCA1, ATP binding cassette transporter A1; ABCG1, ATP transporter G1;LCAT, lecithin:cholesterol acyltransferase; CETP, cholesteryl ester transfer pro-tein; NMR, nuclear magnetic resonance; HDL-P, HDL particles; IDEAL, Decreasein Endpoints through Aggressive Lipid Lowering; VA-HIT, Veterans Adminis-tration HDL Intervention Trial; LpB, lipoprotein B; SR-BI, scavenger receptorclass B type I; CE, cholesteryl ester; PC, phosphatidylcholine; SM, sphingomy-elin; S1P, sphingosine-1-phosphate.

Clinical Chemistry 57:3392–410 (2011) Special Report

392

ture and function of HDL, but it is unclear what impactthese have on CVD risk (11 ). Moreover, developmentof diagnostic and treatment strategies to study and tar-get HDL metabolism must involve not only the abso-lute concentration of HDL cholesterol, but also thefunctional qualities of HDL particles (10, 12 ).

Methods for measurement of HDL subfractions(13 ) as well as compositional and functional assaysmay be superior to HDL cholesterol in predicting cor-onary heart disease (CHD) risk (10, 14 ). Thus, there isa need to propose a new framework to encompass andhighlight the structural, compositional, and functionaldiversity of these particles.

In this special report we discuss the advantages anddisadvantages of current analytical measures of HDLand provide insights into newer research methods thatcharacterize HDL on the basis of its heterogeneousphysicochemical and functional properties. There is anincreasing need to understand, validate, and quantifythe diverse roles of HDL particles in the atheroscleroticprocess to improve diagnosis, prevention, and treat-ment of CVDs (9, 10 ). The information in this report isintended to serve as a foundation to foster improvedunderstanding of the pathophysiology of atherosclero-sis and direct the future course of research and the de-sign of interventions that effectively reduce residualrisk in various patient populations. Finally, we presenta uniform nomenclature for HDL subfractions andpropose a paradigm to define the dynamic process ofHDL metabolism with multiple static laboratorymeasurements. We acknowledge that the variousHDL methods measure different physical and chem-ical properties of HDL and that the use of static mea-sures for dynamic processes has inherent limita-tions, but we also recognize the inconsistency ofexisting nomenclature for HDL subclasses and theneed for a uniform structure that allows clinician-scientists to interrelate the various methods into afunctional construct.

HDL Cholesterol and Cardiovascular Risk

The cholesterol content of HDL is conventionally usedto assess the multifarious antiatherothrombotic andimmune-related functions of HDL particles. The reli-ance on HDL cholesterol in clinical practice partly de-rives from its use as a principal component of theFriedewald equation used to estimate LDL cholesterol(15 ).

HDL cholesterol has been evaluated as a riskmarker in 68 long-term population-based studies in-volving more than 300 000 individuals (16 ). In multi-variate models adjusted for both nonlipid and lipid(triglycerides and non-HDL cholesterol) risk factors,HDL cholesterol is inversely associated with CHD

events. For every 0.39 mmol/L (15 mg/dL) increase inHDL cholesterol concentration, the risk of a CHDevent was reduced by 22% (95% CI, 18%–26%). Lowconcentrations of HDL cholesterol predict CHD mor-tality equally well in nondiabetic and diabetic patients(17 ).

Evidence for the utility of HDL cholesterol as a riskmarker in patients treated with lipid-altering therapyhas been conflicting depending on the extent of covari-ate adjustments. In a metaanalysis involving 90 056 par-ticipants from 14 randomized trials of therapy with st-atins (hydroxy-methyl-glutaryl coenzyme A–reductaseinhibitors), the Cholesterol Treatment Trialists’ Col-laboration investigators reported the proportional ef-fects of various baseline prognostic risk factors on vas-cular events including HDL cholesterol (18 ). The5-year incidence of major CVD events was higheramong individuals with the lowest HDL cholesterolconcentrations. The use of statins reduced the risk ofCHD events by 22% in individuals in the lowest tertileof HDL cholesterol [�0.9 mmol/L (35 mg/dL)] and by21% in individuals in the middle tertile [0.9 –1.1mmol/L (35– 42 mg/dL)] and upper tertile [�1.1mmol/L (42 mg/dL)]. However, participants with thelowest HDL cholesterol concentrations had higher ab-solute risk (22.7%, 18.2%, and 14.2% for the low, mid-dle, and high tertiles, respectively), and thus experi-enced the largest absolute risk reduction.

Similarly, on-trial concentrations of HDL choles-terol are predictive of recurrent CVD events in mostprospective clinical trials (7, 19 ). The increased risk as-sociated with low concentrations of HDL cholesterolpersists even in statin-treated patients with LDL cho-lesterol �1.8 mmol/L (70 mg/dL). However, this con-cept was recently challenged in a metaanalysis of 95trials involving nearly 300 000 individuals, the resultsof which suggested that on-trial HDL cholesterol con-centrations were not significantly related to CHDevents (20 ). Limitations of the study included use ofpooled data rather than individual study participantdata, and lack of consideration of baseline triglycerideconcentrations. Furthermore, the majority of studiesincluded in this metaanalysis showed minimal(�3%) differences in HDL cholesterol concentra-tions between the treatment groups, whereas the an-alytical variation for direct HDL cholesterol assays isfrequently �10% (21 ).

Considering that high-risk individuals are oftentreated with statins, HDL measurements that extendbeyond its cholesterol content may provide more use-ful information for risk stratification in potentiallyhigh-risk individuals and particularly in those patientstreated with lipid-altering therapy.

HDL Nomenclature Special Report

Clinical Chemistry 57:3 (2011) 393

Analytical Limitations of HDL CholesterolMeasurements

The earliest methods for measurement of HDL choles-terol involved preparative ultracentrifugation (22 ) forisolation of HDL with densities between 1.063 and 1.21g/mL. It was not until the advent of chemical-basedprecipitation methods, with reagents such as dextransulfate in the early 1970s, that it became practical tomeasure HDL cholesterol in clinical laboratories. Inthe past 10 years, most laboratories have switched todirect (homogenous) assays that do not involve physi-cal separation of HDL from other lipoproteins. Thereare currently 7 different direct HDL cholesterol assays,which use several different approaches for eithershielding or selectively consuming cholesterol on non-HDL lipoprotein fractions (Table 1). Direct HDL cho-lesterol assays are fully automated and precise and re-quire less labor. Therefore, they have largely replacedolder assays. It remains uncertain whether direct HDLcholesterol assays have clinical utility comparable tothat of chemical-based precipitation methods (23–25 ).In a recent study of 175 individuals with a wide varietyof lipid disorders, none of the 7 current direct assaysmet the minimum total error goal of less than 12%established by the National Cholesterol Education Pro-gram (21 ). Furthermore, inaccurate HDL cholesterolresults from direct assays were found to significantly

compromise the accurate classification of CVD riskbased on estimated LDL cholesterol.

Classification of HDL by PhysicochemicalProperties

ANALYTICAL ULTRACENTRIFUGATION

The earliest method used for quantification of HDLinvolved analytical ultracentrifugation using schlierenoptics. In the late 1940s, Gofman, Lindgren, and col-leagues at Donner Laboratory in Berkeley, California,identified HDL subclasses as a function of size and den-sity based on their ultracentrifugal flotation rate (F1.2)in a high-salt solution (26 ). These studies establishedthat most HDL particles have buoyant density between1.063 and 1.21 g/mL, and this information providedthe basis for standard preparative ultracentrifugal iso-lation of HDL (21, 26 ) (Fig. 1). Moreover, smaller,denser HDL3 (F1.2 0 –3.5) were well distinguished fromlarger, more buoyant HDL2 (F1.2 3.5–9) on the basis ofa distinct shoulder in the optical schlieren profile.Larger HDL (F1.2 9 –20), designated HDL1, repre-sented a relatively minor species in most individuals.Using first principles of physics, we converted the ana-lytical ultracentrifuge schlieren profiles to mass con-centrations of lipoprotein particles. This gold standardmethod was the first to be used in a prospective study todemonstrate an inverse relation of plasma HDL con-centration to CHD risk (27 ). Recently, results of long-term follow-up (29 years) of 1905 men in this studyhave demonstrated that both HDL2 and HDL3 are in-dependently related to CHD risk (28 ).

Table 1. Commercially available directHDL-cholesterol assays.

Precipitation

Heparin-Mn2�

0.46 mmoL (Lipid Research Clinics method)

0.92 mmoL (for EDTA plasma)

Dextran sulfate (50 kDa) Mg�2 (designated comparisonmethod)

Phosphotungstate-Mg2�

Polyethylene glycol (does not precipitate apo E–rich HDL)

Facilitated precipitation

Polymedco (magnetic beads conjugated with dextransulfate-Mg2�)

Direct (homogenous methods)

Denka Seiken (selective elimination)

Kyowa Medex (polyethylene glycol–modifiedenzymes/cylodextrin)

Sekisui Medical (formerly Daiichi) (syntheticpolymer/detergent)

Serotec

Sysmex International Reagents (immunoinhibition)

UMA

Wako Pure Chemical Industries (immunoinhibition)

09

HDL subclass

3 5

1 2 3F1.20

o

20 09 3.5

Mass concentra�on

(AU )

Fig. 1. Analytical ultracentrifugation.

Measurement of HDL by analytical ultracentrifugation. Ma-jor HDL-particle subclasses are distinguished by flotationrate in a salt solution of density 1.2 g/mL (F1.2) and totalmass, represented by the area under the curve (AUC), isdetermined by first principles of physics from schlierenoptics. Initially, 3 major HDL subclasses were identified.HDL1, with the highest flotation rate, is generally notdetectable in substantial concentrations in human plasma.A curve-fitting procedure was later developed to resolve 2subclasses of HDL2 (HDL2a and HDL2b).

Special Report

394 Clinical Chemistry 57:3 (2011)

NONDENATURING GRADIENT GEL ELECTROPHORESIS

Gradient gel electrophoresis in conjunction with au-tomated densitometry was applied in 1981 by Ni-chols and colleagues at Donner Laboratory (29 ) toidentify 5 HDL subspecies separable on the basis ofparticle diameter: HDL3c (7.2 to 7.8 nm), HDL3b(7.8 to 8.2 nm), HDL3a (8.2 to 8.8 nm), HDL2a (8.8to 9.7 nm), and HDL2b (9.7 to 12.9 nm) (Fig. 2).Results of subsequent studies indicated that HDL2b,which is strongly correlated with total HDL choles-terol, was most strongly inversely related to CHDrisk (30 ), and that increased HDL3b was associatedwith an atherogenic lipoprotein phenotype charac-terized by increased triglycerides and small, denseLDL, together with reduced HDL2b (31 ). As de-scribed below, the use of 2-dimensional (2-D) elec-trophoresis has demonstrated that particles corre-sponding to HDL2b are independently and inverselyrelated to CHD risk (32 ).

Density Gradient Fractionation of PlasmaLipoproteins

Precise and reproducible fractionation of the majorHDL particle subpopulations (HDL2b, -2a, -3a, -3b,

and -3c) in human plasma is based on an isopycnicequilibrium method developed by Chapman and col-leagues (33–36 ) (Fig. 3). By use of a gradient tube of aswing-out rotor, the gradient is constructed by consec-utive layering of 4 salt solutions of distinct densitiesthat have been adjusted accurately at the same temper-ature as that of the ultracentrifugal separation(�15° C). The major disadvantage of this method isthe same as that of other ultracentrifugal separations;because lipoproteins are subject to high ionic strengthand centrifugal force (57 � 107g average/min), shearforces are reduced by use of a swing-out rotor.

HDL Diameter Samples

17.0 nmsubclass (nm) #1 #2 #3 #4 Std

12.2 nm12.9

9 5 nm9.7

HDL2b

9.5 nm

8.2 nm

8.8

8.2

HDL2a

HDL3a

7.8HDL3b

7.1 nm7.2

HDL3c

Fig. 2. Nondenaturing gel electrophoresis.

Separation of HDL subclasses from 4 representative plasmasamples by gradient gel electrophoresis. HDL is isolatedfrom plasma by ultracentrifugation at density 1.21 g/mL,and electrophoresed in a 4%–30% nondenaturing gradientgel. After the protein is stained with Coomassie Blue andthe gels are scanned by densitometry, the size distributionis determined by calibration using protein standards (rightlane). This procedure resolves 5 distinct subclasses, al-though the smallest, HDL3c, is generally present at lowconcentrations. Std, standard.

Peak diameter**HDL2b

10.17 nmHDL2a9.14 nm

HDL3a8.75 nm

HDL3b8.52 nm

HDL3c8.47 nm

Fig. 3. Density gradient ultracentrifugation.

Representative electrophoresis profiles and mean particlesizes of HDL subfractions (HDL2b, HDL2a, HDL3a, HDL3band HDL3c) from normolipidemic human plasma separatedby isopycnic, single-spin, density gradient ultracentrifuga-tion [1-dimensional electrophoresis was performed in non-denaturing gradient polyacrylamide gel (4%–20%)]. Hu-man plasma is separated by isopycnic, single-spin, densitygradient ultracentrifugation. The plasma or serum sample(3 mL) adjusted to a density of 1.21 g/mL is layered on acushion of NaCl-KBr solution of density 1.24 g/mL at thebase of the gradient tube; the discontinuous gradient isthen completed by layering successive density solutions of1.063, 1.019, and 1.006 g/mL onto the latter. The proce-dure involves a single ultracentrifugal step, allows almostquantitative recovery of highly resolved HDL fractions ofdefined hydrated density and physicochemical properties,avoids major contamination with plasma proteins, andfacilitates HDL isolation in a nondenatured, nonoxidizedstate. Gradients are fractionated with a precision pipettefrom the meniscus downwards, thereby avoiding contam-ination with plasma proteins �1.25 g/mL present in theresidue at the base of the tube. Peak diameter was deter-mined at the maximum absorption intensity of each bandby using Kodak 1D software filters following staining withCoomassie Brilliant Blue. **Size by negative stain electronmicroscopy provided smaller estimates (HDL2b � HDL2a,mean diameter 9.6 nm and range 10.8–7.2 nm; HDL3a �HDL3b � HDL3c, mean diameter 7.3 nm and range 9.0–5.4 nm) reflecting the nonhydrated state.

HDL Nomenclature Special Report

Clinical Chemistry 57:3 (2011) 395

VERTICAL ANALYTICAL PROFILE

Vertical auto profile (VAP) is another HDL subfrac-tionation method based on ultracentrifugation (37 ).Unlike most other ultracentrifugation methods, VAP isdone in a vertical rotor, which makes the method rela-tively fast and more practical for the analysis of routineclinical specimens. With regard to HDL, VAP measuresthe cholesterol content of its 2 major density subfrac-tions, namely HDL2 and HDL3 (38 ).

VAP is relatively precise, with intraassay CVs forlipoprotein subfractions that range from 4% to 10%

(39 ). Specific limitations of ultracentrifugationmethods including the VAP technique are describedin online Appendix 1 in the Data Supplement thataccompanies the online version of this article athttp://www.clinchem.org/content/vol57/issue3.

Limited studies have been done to compare theVAP method to other lipoprotein subfraction proce-dures, but so far the results of these studies have shownrelatively poor agreement (40 ). This is perhaps not sur-prising because of lack of standardization of the differ-ent fractionation methods, as well as the fact that thevarious lipoprotein subfractionation methods arebased on different physical and chemical properties ofHDL.

2-D GEL ELECTROPHORESIS

HDL can be separated on the basis of size and charge(Figs. 4 and 5) (13, 41 ). The concentrations of theseparticles are expressed in milligrams per liter of apoli-poprotein A-I (apo AI), and as a percentage of totalplasma apo AI concentration. Five major HDL parti-cles have been identified: (a) very small discoidal pre-cursor HDL of pre-� mobility (pre-�-1 HDL, diameterabout 5.6 nm), which contains apo AI and phospho-

Fig. 4. 2-D gel electrophoresis.

The apo AI–containing HDL subpopulation profiles of aCHD patient (A) and a healthy individual (control) (B), witha schematic diagram of all of the apo AI–containing HDLparticles shown on the right (C). Below panel (A) is a plotof a densitometric scan across the �-migrating HDL-particleregion indicating the presence of 4 �-migrating HDL par-ticles ranging in mean particle diameter from very large�-1 HDL (11.0-nm diameter) to very small pre-�-1 HDL(5.6-nm diameter). In the schematic diagram �-migratingapo AI–containing particles in the �-2 region (9.2-nmdiameter) and in the �-3 region (8.1-nm diameter) containboth apo AI and apo AII (more heavily shaded), whereas allother particles containing apo AI, including small �-4 HDL(7.4-nm diameter), do not contain appreciable amounts ofapo AII (less heavily shaded). The asterisk marks the serumalbumin or � front. Based on their composition, very smallpre-�-1 HDL and small �-4 HDL are discoidal particles thatdo not contain cholesteryl ester or triglyceride, whereasmedium, large, and very large �-3, -2, and -1 HDL arespherical and contain cholesteryl ester and triglyceride intheir cores. CHD patients in general in the untreated statetend to have significant decreases in the levels of apo AI invery large and large �-migrating HDL and modest increasesin apo AI in very small pre-�-1 HDL and small �-4 HDL. In(C), 1, 2, 3, and 4 refer to �-1, -2, -3, and -4, respectively.Asztalos et al. (181 ).

Fig. 5. 2-D gel electrophoresis patterns after apo AIimmunoblotting.

The 2-D gel electrophoresis patterns after apo AI immu-noblotting observed for whole plasma are shown in theleft panel, for lipoproteins of density (d) �1.125 g/mL asseparated by ultracentrifugation are shown in the centerpanel, and for lipoproteins of density 1.125–1.24 g/mLare shown in the right panel. These data indicate thatapo AI– containing HDL of density �1.125 g/dL arecomprised mainly of very large and large �-migratingHDL, whereas apo AI– containing HDL particles of den-sity 1.125–1.24 g/mL contain mainly medium and small�-migrating HDL and very small pre-�-1 HDL. Asztaloset al. (181 ).

Special Report

396 Clinical Chemistry 57:3 (2011)

lipid; (b) very small discoidal HDL of � mobility (�-4HDL, diameter about 7.4 nm), which contains apo AI,phospholipid, and free cholesterol; (c) small sphericalHDL of � mobility (�-3 HDL, diameter about 8.0 nm),which contains apo AI, apo AII, phospholipid, freecholesterol, cholesteryl ester, and triglyceride; (d)medium-sized spherical HDL of � mobility (�-2 HDL,diameter about 9.2 nm), which contains the same con-stituents as � 3 HDL; and (e) large spherical HDL of �mobility (�-1 HDL), which contains the same constit-uents as �-3 and �-2 HDL, except for the near absenceof apo AII (Fig. 4). Adjacent to the � particles are pre-�particles that are of similar size, but present in loweramounts and do not contain apo AII. In addition, thereare large pre-� migrating HDL known as pre-�-2 HDL(42 ).

Pre-�-1 HDL particles are most efficient in inter-acting with the ATP binding cassette transporter A1(ABCA1) to promote cholesterol efflux from cells,whereas large �-1 HDL are the most efficient in inter-acting with the liver scavenger receptor B1 for deliveryof cholesterol to the liver (43, 44 ). Intermediate-sized�-3 HDL are the most efficient in interacting with theATP transporter G1 (ABCG1) to promote cellular cho-lesterol efflux onto spherical HDL particles containingboth apo AI and apo AII (44 ). Delipidated HDL or apoAIMilano complexed with phospholipids, which havebeen reported to promote regression of coronary ath-erosclerosis when delivered by infusion, consist of pre-�-1 HDL particles (45 ). There are other HDL particlescontaining apo E without apo AI (very large pre-� mi-grating HDL) and small HDL containing apo AIVwithout apo AI (43 ). The functions of these latter par-ticles have not been well defined.

2-D gel electrophoresis of plasma followed by apoAI immunoblotting allows for the accurate diagnosis ofdisorders of HDL metabolism. apo AI deficiency ischaracterized by the absence of apo AI– containingHDL particles, and patients with apo AI deficiency of-ten develop xanthomas and early-onset CHD (46 ). apoAI is nearly absent in Tangier disease, which is charac-terized by lack of functional ABCA1 transporters andcholesteryl ester deposition in macrophages through-out the body. These patients have only pre-�-1HDL, and they usually develop premature CHD (47 ).Familial lecithin:cholesterol acyltransferase (LCAT)-deficient patients have only pre-�-1 and �-4 HDL par-ticles, cannot esterify their cholesterol, and can developsevere corneal opacification, increased LDL, and renalfailure (41 ). Lipoprotein lipase– deficient patients havemarked hypertriglyceridemia that places them at highrisk for pancreatitis. These patients also have low con-centrations of HDL cholesterol that is carried only inpre-�-1 and �-4 HDL particles (48 ). Hepatic lipase–deficient patients have increased remnant lipoproteins,

a decrease in �-2 HDL, and increased risk for prema-ture CHD (48 ). Cholesteryl ester transfer protein(CETP)-deficient patients have very large � HDL thatcontains apo AI, apo AII, and apo E (49 ). The presenceof apo E on these large HDL particles may be importantto drive cholesterol efflux by the ABCG1 transporter.

When the concentration of apo AI in �-1 HDL isless than 140 mg/L, the individual is at increased risk ofdeveloping CHD (32 ). CHD patients also often havesmall discoidal HDL particles and decreased large �-1and �-2 HDL particles. Concentrations of these parti-cles are superior to HDL cholesterol concentrations foruse in CHD risk prediction (50, 51 ).

Large �-1 particles increase with weight loss, nia-cin, certain statins (atorvastatin, rosuvastatin), andCETP inhibitors (52–57 ). Increasing the concentra-tions of apo AI in �-1 HDL to �200 mg/L (0.52mmol/L) with a simvastatin/niacin combination hasbeen associated with lack of progression and insome individuals with regression of coronaryatherosclerosis (51 ).

HDL Particle Concentration

NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY

Unlike other methods for analyzing HDL particles, nu-clear magnetic resonance (NMR) spectroscopy doesnot require a physical separation step because the pro-tons (the nuclei of hydrogen atoms) within lipoproteinparticles of different sizes have a natural magnetic dis-tinctness arising from their unique physical structure(58 ). As a result, lipoprotein particles of different sizesin unfractionated plasma or serum give rise to distin-guishable lipid NMR signals that have characteristicallydifferent frequencies (Fig. 6, left panel) (59, 60 ). TheNMR signal frequencies (chemical shifts) of HDL sub-fractions, compared to LDL and VLDL subfractions,are particularly well differentiated (Fig. 6, right panel).

The lipoprotein NMR signals employed for li-poprotein quantification are those from the terminallipid methyl group protons, because they are unre-sponsive to, and therefore unaffected by, fatty acid andother chemical compositional differences (60 ). Fur-thermore, to a close approximation the number ofmethyl protons in a lipoprotein particle of given diam-eter is constant even in the face of significant variationsin core cholesterol ester and triglyceride content. Theseproperties render the detected subclass methyl-signalamplitudes directly proportional to subclass particlenumber and enable NMR-derived concentrations ofHDL to be given in particle number units (�mol/L)(60 ). Although NMR analysis provides a new and po-tentially clinically advantageous way to quantify HDLand its subfractions, the NMR lipid methyl signal is

HDL Nomenclature Special Report

Clinical Chemistry 57:3 (2011) 397

inherently unable to provide HDL chemical composi-tional information.

The current NMR method models the plasma sig-nal as the sum of the contributing signals of 26 sub-populations of HDL as well as 47 subpopulations ofLDL, VLDL, and chylomicrons. Given the limited mea-surement precision of the derived concentrations ofeach of the many subpopulations, they are grouped forroutine reporting purposes into “large,” “medium,”and “small” subclass categories. For research studies, itis not difficult to produce alternative groupings of the26 HDL subpopulations to make the reported subfrac-tions more similar to those produced by other analyticmethods such as density gradient ultracentrifugationand gradient gel electrophoresis.

Investigations are ongoing to establish CVD rela-tions with NMR-determined concentrations of HDLparticles, HDL subclasses (large HDL particles, 9.4 –14nm; medium HDL particles, 8.2–9.4 nm; and smallHDL particles, 7.3– 8.2 nm), and HDL particle size.Among published studies are those documenting as-

sociations with age and sex (61, 62 ), longevity (63 ),insulin resistance and diabetes (64 – 67 ), CHD(62, 68 –74 ), and changes brought about by exercise(75 ) and treatment with various drugs (76 – 84 ). Animportant consideration in interpreting the clinicalsignificance of observed univariate disease associa-tions with individual HDL subclasses or HDL size isthe confounding that arises from the strong inversecorrelation between large and small HDL subclassesand the even stronger inverse associations of largeHDL particles (and HDL size) with total (and small)LDL particle concentrations (60, 65, 74 ). Withoutconducting statistical analyses that address the con-founding caused by these correlations, misleadingconclusions may be reached about the clinical im-portance and potential functional differences amongHDL subclasses (62, 73, 74 ).

ION MOBILITY

Ion mobility, a gas-phase differential electrophoresismacromolecular mobility-based method, was devel-

HDL

LDL

VLDL/Chylos

0 20 40 60 80 100 120 140Lipoprotein diameter (nm)

Rel

ativ

e N

MR

freq

uenc

y (H

z)

0

4

8

12

16

20

24

NM

R c

hem

ical

shi

ft (p

pm)

0.78

0.79

0.80

0.81

0.82

0.83

0.84

0.82 0.78 0.74

Methyl chemical shift (ppm)

Isolated HDL subclasses11.5 nm

9.4 nm

8.5 nm

8.0 nm

7.5 nm

Fig. 6. NMR spectroscopy.

Relationship of lipoprotein subclass diameter to lipid methyl group NMR chemical shift and frequency (left panel) and methylsignal line shape and chemical shift of 5 isolated HDL subclasses of differing diameter (right panel). The natural magneticdistinctness of lipoprotein particles of different size makes it possible in theory to use any NMR instrument in any laboratoryfor lipoprotein analysis, but in practice such analysis requires dedicated instrumentation. The subclass NMR signals highlyoverlap, making it necessary to computationally “deconvolute” the plasma NMR signal envelope to extract the amplitudes ofthe subclass signals that are used to calculate subclass concentrations. Accurate and reproducible deconvolution is possible onlyif the NMR conditions (such as magnetic field strength and temperature) used to generate the subpopulation reference signallibrary exactly match the conditions used subsequently to measure (in approximately 1 min) each patient plasma sample. TheNMR LipoProfile-3 method currently employed by LipoScience models the plasma signal as the sum of the contributing signalsof 26 subpopulations of HDL as well as 47 subpopulations of LDL, VLDL, and chylomicrons (Chylos). Given the limitedmeasurement precision of the derived concentrations of each of the many subpopulations, they are grouped for routinereporting purposes into “large,” “medium,” and “small” subclass categories.

Special Report

398 Clinical Chemistry 57:3 (2011)

oped by Benner and colleagues at Lawrence BerkeleyNational Laboratory (85 ). In this high-throughputprocedure, lipoprotein size is determined by first prin-ciples of physics, and particles are counted directly afterequalization of charge and separation by time of flightthrough a voltage gradient. Albumin contamination ofthe HDL size region is first reduced by incubation withblue dextran and a short ultracentrifugation in the ab-sence of salt. In its current configuration, the method isdesigned to separate HDL2b from smaller HDL; meth-ods are in progress to resolve and measure HDL2a andthe HDL3 subspecies. This method has recently dem-onstrated that the large HDL2b subspecies is stronglyinversely correlated with coronary disease risk in theprospective Malmo Diet and Cancer Study (86 ). Theassociation of these large HDL particles with CHD riskwas related to their inclusion in 2 independent princi-pal components determined from ion mobility mea-surements of all lipoprotein fractions. One of these cor-responds to the atherogenic lipoprotein phenotype,which includes increased concentrations of triglycer-ides and smaller LDL particles, and the second includessmaller HDL particles. Results of genetic analyses per-formed in this study indicated that these componentshave differing underlying determinants, and this mayindicate 2 independent mechanisms for the cardiopro-tective effects of HDL.

HDL Particle Heterogeneity Viewed through Lipidand Protein Composition: apo AI andCardiovascular Risk

apo AI is the major HDL protein (87 ) and is synthe-sized by both hepatic and intestinal cells (88 ). apo AIhas been considered a more precise biomarker thanHDL cholesterol on the basis of its functional role inmediating cholesterol mobilization via the ABCA1transporter in peripheral cells, including arterial mac-rophages. Indeed, early studies suggested that apo AImeasurements were superior to HDL cholesterol as arisk marker (89, 90 ). Subsequently, these 2 measures ofHDL, HDL cholesterol and apo AI, have been directlycompared in 2 large cohorts (74, 91 ). In the EuropeanProspective Investigation into Cancer and Nutrition–Norfolk study, which included 2349 individuals, thenon–lipid-adjusted risk of a major CHD event per1-SD change was 0.78 (0.70 – 0.87) for HDL cholesteroland 0.79 (0.71– 0.87) for apo AI (74 ). A recent analysisof the Women’s Health Study showed that the magni-tude of risk for a major cardiovascular event was higherfor low levels of HDL cholesterol than for apo AI (91 ).

Among high-risk individuals treated with statintherapy, on-trial measures of apo AI provide incre-mental information on CHD risk beyond that obtainedwith HDL cholesterol (92 ), whereas in the stable CHD

patients enrolled in the Incremental Decrease in End-points through Aggressive Lipid Lowering (IDEAL)study, these measures provide equivalent prognosticdata (74 ). In the Air Force-Texas Coronary Atheroscle-rosis Prevention Study, low on-trial concentrations ofapo AI at 1 year were predictive of major CHD events,whereas no significant predictive value was seen foron-trial HDL cholesterol concentration (92 ). In con-trast, there were no differences between HDL choles-terol and apo AI concentrations with regard to risk ofrecurrent events in statin-treated CHD patients en-rolled in the IDEAL study (74 ) or fibrate-treated CHDpatients enrolled in the Veterans Administration HDLIntervention Trial (VA-HIT) (71 ).

apo AI Measurements

Because apo AI is an abundant serum protein, it is rel-atively easy to measure by either nephelometric orturbidometric methods that are available on moststandard clinical chemistry analyzers (93 ). Usually,nonionic detergents are added to the assay buffer todisrupt HDL and expose apo AI antigenic sites, whichameliorates background problems from turbid speci-mens. Oxidized forms of apo AI (94 –96 ) have alsobeen evaluated as potential CHD risk biomarkers andmay stimulate a reconsideration of the value of apo AIin cardiovascular risk assessment.

Classification of HDL by ApolipoproteinComposition of Lipoprotein Particles

Plasma lipoproteins can be separated and classified onthe basis of their apolipoprotein composition. Alaupo-vic and colleagues used apolipoprotein composition asa basis for separating human plasma lipoproteins intoparticles, which included lipoprotein B (LpB) (LpB,LpB:C, LpB:C:E), LpA (LpAI, LpAII, LpAI:AII), LpC(LpCI:CII:CIII), LpE, and LpD (97, 98 ).

apo AI AND apo AI:AII PARTICLES

LpAI and LpAI:AII are the major HDL lipoproteins,containing approximately 35% and 65% of plasma apoAI, respectively (99 ). LpAI is initially secreted as alipid-poor apo AI:phospholipid complex, which inter-acts with the ABCA1 transporter and facilitates choles-terol efflux resulting in the formation of pre-� HDL(100 ). The cholesterol in pre-� HDL is esterified tocholesteryl esters by LCAT, with conversion of pre-�HDL– containing to � HDL– containing LpAI particles(100, 101 ). Hepatic secreted apo AII associates withLpAI to form LpAI:AII. The role of apo AII in HDLmetabolism has not been definitively established, al-though apo AII has been reported to decrease HDLremodeling (102 ) and reduce cholesterol uptake by he-

HDL Nomenclature Special Report

Clinical Chemistry 57:3 (2011) 399

patic scavenger receptor class B type I (SR-BI) (103 ). �HDL– containing LpAI and LpAI:AII drive efflux ofcholesterol after interaction with the ABCG1 trans-porter (104 –106 ). Thus, a dual pathway for efflux ofcholesterol from cholesterol-loaded cells involveslipid-poor AI particles interacting with the ABCA1transporter, and larger LpAI/LpAI:AII particles inter-acting with the ABCG1 transporter (107–111 ). Inplasma both LpAI and LpAI:AII are heterogeneous andcan be separated into subfractions on the basis of lipidcomposition, density, size, and charge.

The cardioprotective roles of LpAI and LpAI:AIIhave been controversial. In an initial report LpAI butnot LpAI:AII was able to drive cholesterol efflux fromOb1771 adipocytes in culture, suggesting that LpAI butnot LpAI:AII was antiatherogenic (110 ). Results ofsubsequent clinical studies in which LpAI and apoAI:AII were evaluated in CHD patients have revealed adecrease in HDL cholesterol and LpAI:AII (110 ) and aselective decrease in LpAI in individuals with HDLcholesterol �40 mg/dL (1.03 mmol/L) (112 ); In theEtude Cas-Temoins sur l’Infarctus du Myocarde trial,LpAI was decreased in CHD patients in Northern Ire-land; however, both LpAI and LpAI:AII were decreasedin patients in France (113 ). In the Prospective Epide-miologic Study of Myocardial Infarction study of 8784individuals from France and Northern Ireland, regres-sion logistic analyses showed that apo AI was a strongerpredictor for the risk of CHD than HDL cholesterol,LpAI, and LpAI;AII (114 ). Quantification of LpAI andLpAI:AII in the Framingham Offspring study andVA-HIT clinical trial did not differentiate a subset ofindividuals with increased risk of CHD after adjust-ment for lipid and non-lipid CHD risk factors (32, 50 ).The variability in the results observed in analyses ofLpAI and LpAI:AII in the various clinical studies mayreflect potential heterogeneity in HDL particles in dif-ferent patient groups as well as the diverse methodsused to quantify HDL subclasses. A general conclusionresulting from these studies is that increased HDLcholesterol concentrations involving an increase inthe large HDL subclasses containing both LpAI andLpAI:AII is associated with decreased CHD risk,whereas the reduction in lipid-poor LpAI and pre-�HDL is associated with increased CHD risk (115 ).

In healthy individuals, LpAI is catabolized at afaster rate than LpAI:AII (116 ). The major sites of ca-tabolism of the protein components of LpAI andLpAI:AII are the liver and kidney, and the majority ofHDL cholesterol is transported to the liver. Kineticmodels incorporating the different rates of catabolismof LpAI and LpAI:AII have been developed (117, 118 ).The differential rates of metabolism of LpAI and LpAI:AII have been proposed to be related to the decreasedability of apo AI– containing lipoproteins to reassociate

with HDL particles following cholesteryl ester deliveryto the liver via the SR-BI receptors. Lipid-poor apo AIparticles, in contrast to apo AII– containing HDL par-ticles, are rapidly catabolized by the kidney, leading toan increased fractional catabolic rate (119 ).

Genetic absence of plasma LpAI resulting from amolecular defect in apo AI results in increased CVD(120 –122 ), whereas a genetic defect in apo AII result-ing in LpAI:AII deficiency was not associated with amajor clinical phenotype (11, 123 ). Increased catabo-lism of LpAI and LpAI:AII leading to low HDL choles-terol concentrations is characteristic of Tangier diseaseand LCAT deficiency. Genetic deficiency of the ABCA1transporter (Tangier disease) is associated with de-creased cholesterol efflux and poor lipidation of pre-�HDL, resulting in accelerated LpAI catabolism and in-creased CVD (124 ). In contrast, in LCAT deficiencythere is effective cholesterol efflux from cholesterol-loaded macrophages followed by defective maturationof pre-� HDL to � HDL; accelerated catabolism, pri-marily of LpAI:AII as well as LpAI; and renal disease,but no increased risk of CVD (125, 126 ). A unique li-poprotein particle, LpAI:AII:E, with attenuated catab-olism relative to LpAI, is present in patients with CETPdeficiency and markedly increased plasma HDL cho-lesterol concentrations (127, 128 ). Decreased catabo-lism of LpAI and LpAI:AII with increased HDL choles-terol concentrations has also been reported in patientstreated with CETP inhibitors (52 ). Decreased plasmalevels of LpAI and LpAI:AII are seen in hypertriglyeri-demic patients because of increased catabolism oftriglyceride-enriched LpAI and LpAI:AII (129, 130 ).

Statin drugs are associated with a modest 5%–7%increase in HDL cholesterol concentrations due tocomplex alterations in LpAI and LpAI:AII metabolism,including increased synthesis and decreased catabo-lism of apo AI as well as a decrease in CETP activity(131–133 ). In an analysis of the Voyager database thatincluded 37 randomized clinical trials involving 32 258patients, the percentage increase in HDL cholesteroland apo AI associated with the administration of ator-vastatin, simvastatin, and rosuvastatin were similar(134 ). In human clinical studies fenofibrate adminis-tration resulted in an increased synthesis of apo AIIwith a minimal increase in apo AI synthesis; overall,increases in LpAI:AII predominated (131, 133, 135 ).Recent human kinetic studies with niacin have demon-strated both an increased synthesis and catabolism ofapo AI and apo AII resulting in the preferential forma-tion and accumulation of large LpAI particles and in-creased HDL cholesterol concentrations (53 ).

apo E

apo E is the most avid ligand for the LDL receptor, andas such drives catabolism and clearance of apo B–

Special Report

400 Clinical Chemistry 57:3 (2011)

containing lipoproteins (136, 137 ). apo E is also aprominent HDL protein with unique functions in theHDL compartment (138, 139 ). For example, swine anddogs fed a high-fat diet accumulate large apo E–richHDL that can deliver cholesterol to the liver directly viainteraction with the LDL receptor (140, 141 ). In thepresence of apo E, HDL particles undergo a core expan-sion due to their enhanced capacity to carry cholesterol(142 ). In addition, apo E interacts with ABCA1 to ex-tract cellular cholesterol out of the cell, and drives theformation of larger HDL-sized particles from macro-phage foam cells (143, 144 ). Because atheromatousplaques are only partially permeable to plasma solutes,such as apo AI, but rich in locally secreted proteins,such as apo E, arterial macrophages may represent cellsin a unique microenvironment in which cholesterolefflux is directed more toward apo E– containing par-ticles rather than toward classical apo AI– containingparticles (145, 146 ).

In humans, apo E HDL concentrations are lowerthan in animals lacking CETP and vary with fastingstate and apo E phenotype (147–149 ). Interestingly,both CETP-deficient patients (149, 150 ) and individu-als treated with a CETP inhibitor (151 ) have increasedapo E HDL concentrations (48, 150 ). The apo E–richHDL of CETP-deficient individuals appears to be astrong acceptor of ABCG1-derived cholesterol fromloaded macrophages. Finally, apo E inhibits the dis-placement of hepatic lipase from the endothelial sur-face (152 ), thus reducing hepatic lipase-mediated tri-glyceride hydrolysis of HDL and the affinity of HDL forits docking receptor, SR-BI (153 ). These observationssuggest a scenario in which, in conditions of dimin-ished CETP activity, the HDL uses apo E for core ex-pansion and for direct hepatic delivery of lipid cargovia the LDL receptor (154 ). However, it remains un-clear whether HDL particles present in CETP-deficientpatients or generated by use of CETP inhibitors areremoved by the LDL receptor, SR-BI, or a combinationof both receptors.

The role of apo E HDL in atherosclerosis is difficult toassess, mostly because of the dominant effects of apo E onwhole-body cholesterol metabolism and lipoprotein dis-position (155). In experimental atherosclerosis, apo E isa potent antiatherogenic agent, not exclusively becauseof its effects on plasma lipids. Small amounts ofmacrophage-derived apo E completely correct both thedyslipidemia and the susceptibility to atherosclerosis inthe apo E–deficient mouse model (156). More impor-tantly, introduction of apo E in the plasma or vascular wallin quantities that are insufficient to change plasma lipidconcentrations still produces significant vascular protec-tion, thereby suggesting local effects within the atheroma(157). In contrast, apo E–deficient patients have not beenreported to manifest early onset of accelerated atheroscle-

rosis (158–160). In support of the contention that apo E ishighly expressed in the atheroma, a recent evaluation ofHDL composition by proteomics analysis has shown apoE enrichment in the HDL3 particles of individuals withCHD (161). This result may provide impetus for the de-velopment of methods aiming at validating HDL apo E asa biomarker to predict the presence of atheroma.

apo M HDL

Genetic manipulation studies in mice indicate that apo Mplays an important role in the remodeling of plasma HDL,formation of pre-� HDL, and reverse cholesterol trans-port and it is a potent antiatherogenic protein (162, 163).apo M is a minor apolipoprotein found in approximately5% of total HDL (see section on Proteomics below) andapproximately 2% of LDL. apo M is positively correlatedwith HDL and LDL cholesterol in both healthy individu-als and CVD patients (164). In 2 case control studies,however, no significant differences were observed inplasma apo M concentrations in apparently healthy indi-viduals and CHD patients (165).

apo B HDL

apo B peptides have been found during shotgun pro-teomics analyses of isolated human HDL (161, 166 ),but their presence has been dismissed as due to eithercontaminating LDL or the presence of lipoprotein (a),the hydrated density of which overlaps that of HDL.Recent studies focused on hepatic microsomal triglyc-eride transfer protein (MTP) activity in mouse modelshave shown that the phospholipid transfer activity ofMTP helps in assembly and secretion of VLDL- andHDL-sized particles containing apo B100 and apo B48(167 ). Low concentrations of these particles are se-creted and may be detected in the plasma.

PROPOSED NOMENCLATURE

As discussed above, use of different techniques andprocedures has led to different terms in defining HDLspecies. To provide guidelines for future studies and tocompare and contrast published data obtained by useof different methods, we propose a new HDL nomen-clature based on density and size of the particles (Table2). In addition, we compare these terms with other des-ignations available in the literature. In this nomencla-ture, HDL particles are termed very large, large, me-dium, small, and very small.

Proteomics and Lipidomics: An Integrated View ofHDL Biology

PROTEOMICS

The advent of the wider availability of mass spectro-metric technologies, and their applicability to analysisof multicomponent protein mixtures, has heralded a

HDL Nomenclature Special Report

Clinical Chemistry 57:3 (2011) 401

surge of interest in the proteome of human HDL par-ticles in health and disease.

Several factors are of central importance whenstudies of the HDL proteome are undertaken, and theseinclude the nature of the starting biological materialand its conservation, the method used for HDL particleseparation and purification, and the type of mass spec-trometric analysis applied. No systematic study of theimpact of such factors on HDL isolation by differentmethods, and thus potentially on the HDL proteome,has been undertaken to date.

The working criterion used to define the HDLfraction under study is a key determinant of the HDLproteome. The choice of isolation/fractionation proce-dures is listed in Table 3; the precise nature of the HDL

isolated by any of these techniques requires rigorousanalysis prior to initiation of proteomic studies. In-deed, HDL isolated by fast-performance liquid chro-matography is heavily contaminated by high molecularweight plasma proteins that coelute with HDL (168 ).Until now, ultracentrifugation has been the predomi-nant method used for HDL isolation to study the HDLproteome.

Once HDL is purified, the mass spectrometrictechnologies that have been employed to define theHDL proteome include SELDI-TOF, MALDI-TOF,shotgun and nano–liquid chromatography electros-pray ionization mass spectrometry, and most re-cently, a shotgun approach involving linear ion trapquadrupole–Fourier transform ion cyclotron reso-nance mass spectrometry with a nanoelectron spraysource. To different degrees, difficulty in quantifyingproteins (as tryptic peptides), particularly those in lowabundance, is a limitation of all of these technologies.

In one of the first comprehensive analyses of theHDL proteome, Vaisar et al. (161 ) identified some 50protein components in human HDL3 isolated by ultra-centrifugation. The biological activities of these pro-teins suggested that HDL contributes not only to lipidmetabolism and cholesterol homeostasis, but also tocomplement regulation, the acute-phase response, andinhibition of proteolytic enzymes. Several other reportshave confirmed the presence in HDL of multiple apo-lipoproteins (AI, AII, AIV, B, (a), CI, CII, CIII, CIV, D,E, F, H, J, L1, and M), in addition to �1-antitrypsininhibitor; albumin; complement C3 and C4; fibrino-gen; haptoglobin-related protein; paraoxonase 1 and 3;

Table 2. Classification of HDL by physical properties.a

Proposed termVery large HDL

(HDL-VL)Large HDL

(HDL-L)Medium HDL

(HDL-M)Small HDL

(HDL-S)Very small HDL

(HDL-VS)

Density range, g/mL 1.063–1.087 1.088–1.110 1.110–1.129 1.129–1.154 1.154–1.21

Size range, nm 12.9–9.7 9.7–8.8 8.8–8.2 8.2–7.8 7.8–7.2

Density gradient ultracentrifugation HDL2b HDL2a HDL3a HDL3b HDL3c

Density range, g/mL 1.063–1.087 1.088–1.110 1.110–1.129 1.129–1.154 1.154–1.170

Gradient gel electrophoresis HDL2b HDL2a HDL3a HDL3b HDL3c

Size range, nm 12.9–9.7 9.7–8.8 8.8–8.2 8.2–7.8 7.8–7.2

2-D gel electrophoresis �-1 �-2 �-3 �-4 Pre-�-1 HDL

Size range, nm 11.2–10.8 9.4–9.0 8.5–7.5 7.5–7.0 6.0–5.0

NMR Large HDL-Pb Medium HDL-P Small HDL-P

Size range, nm 12.9–9.7 9.7–8.8 8.8–8.2 8.2–7.8 7.8–7.2

Ion mobility HDL2b HDL2a and HDL3

Size range, nm 14.5–10.5 10.5–7.65

a One-dimensional electrophoresis was performed in nondenaturing gradient polyacrylamide gel (4%–20%).b HDL-P, HDL particle.

Table 3. Preparative HDL isolation/fractionationtechniques.

• Flotational ultracentrifugation

• Zonal ultracentrifugation

• Density gradient ultracentrifugtion (isopynic, NaCl/KBr; D2O,sucrose)

• Precipitation

• Size exclusion chromatography

• Ultracentrifugation/HDL density 1.063–1.21

• Fast protein liquid chromatography/whole plasma

• Ion exchange chromatography

• Immunoaffinity chromatography

• Electrophoresis (2-D gel electrophoresis, etc.)

Special Report

402 Clinical Chemistry 57:3 (2011)

serum amyloid A1, A2, and A4; and transthyretin [datasummarized in (169 )].

Intriguingly, the plasma abundance of most ofthese proteins is insufficient to allow 1 copy per HDLparticle, thereby suggesting that specific proteins maybe bound to distinct particle species that are differen-tially distributed across the HDL particle spectrum. Onthis basis, it was plausible to expect that the multiplebiological functions of HDL may be mediated by dis-tinct particle subspecies defined by specific cluster(s) ofbound proteins, and that such protein clusters cofrac-tionate upon isolation of HDL subpopulations. Asan initial step toward assessment of this hypothesis,plasma HDL from normolipidemic individuals wassubfractionated by isopycnic density gradient ultracen-trifugation into 5 subfractions (Fig. 3); the composi-tion of their proteome was then evaluated by tandemmass spectrometry technology (166 ).

Five distinct patterns of distribution of individualprotein components were observed across the HDLdensity subfractions; the most interesting of these iden-tified small dense HDL (HDL3c) as a particle subpopu-lation in which 7 proteins occurred predominantly:apo J, apo L1, apo F, paraoxonase 1/3, phospholipidtransfer protein, and platelet-activating factor acetyl-hydrolase (also termed lipoprotein-associated phos-pholipase A2). The HDL3c proteome also containedapo AI; apo AII; apo D; apo M; serum amyloids A1, A2,and A4; apo CI; apo CII; and apo E.

The unique proteome of HDL3c has functionalimplications because this particle species exhibited thegreatest potency among HDL subpopulations to pro-tect LDL against oxidation. Such activity was highlycorrelated with the presence of apo J, apo M, serumamyloid A4, apo D, apo L1, and paraoxonases 1/3 inHDL3c.

These data should not be interpreted to indicatethat all of the proteins detected in HDL3c are presenton the same lipoprotein particle; indeed, the isolationof a unique particle containing the trypanosome lyticfactor apo L1, plus apo AI and haptoglobin-relatedprotein in the HDL3 density range suggests that this iscertainly not the case (169 ), and that the HDL3c frac-tion consists of several species of HDL particles withdistinct proteomes. On the basis of these findings, itmay be concluded that: (a) the detected protein clus-ters described herein are potentially indicative of dis-tinct subspecies of HDL particles that display specificbiological function(s); (b) the proteomic analyses ofdefined HDL subspecies isolated by isopycnic densitygradient ultracentrifugation from normolipidemicplasma samples have identified small dense HDL3c as adistinct particle subset(s); and (c) specific lipid andprotein components of HDL3c endow them with po-tent antioxidative activities. Finally, these data support

the concept that HDL may serve as a platform for theassembly of certain protein components that performspecific function(s), and that (apolipo)proteins mayform the basis for functional heterogeneity of HDL.

It is of special interest to know whether the pro-teome of HDL may be altered in metabolic diseasescharacterized by dyslipidemia and increased CVD risk.If this were the case, then specific proteins could beused as biomarkers of altered HDL function. Indeed, itis now established that several of the major biological,antiatherogenic activities of HDL are attenuated intype 2 diabetes and metabolic syndrome, each of whichis associated with high CVD risk (12 ). Furthermore,under conditions of acute inflammation, HDL parti-cles are enriched in serum amyloid A, resulting in de-fective antiinflammatory activity of HDL (12 ).

Vaisar et al (161 ) and Greene et al (170 ) reportedthe first studies to detect modification of the HDL pro-teome in patients displaying incident CHD, and themost consistent finding was an increment of 150% inapo E content. This alteration in the proteome was nor-malized by combined treatment with a statin and nia-cin. These studies open new horizons not only foridentification of protein biomarkers of altered HDLmetabolism and function but also for well-targetedpharmacotherapies to correct them.

In summary, the precise nature of the HDL pro-teome critically depends on the method of HDL isola-tion and purification, and equally on the mass spectro-metric technology employed for protein analysis andtryptic peptide quantification; structural and func-tional analysis of HDL particle subspecies may provemore informative than analyses of traditional totalHDL; agreement on standardization of methods forHDL isolation from human plasma is especially impor-tant to define the key characteristics of HDL particles.

LIPIDOMICS

When HDL particles are evaluated for their content ofcholesteryl ester and phosphatidylcholine (PC), cho-lesteryl linoleate predominates among cholesteryl es-ter, whereas the 18:2/16:0, 18:2/18:0, and 20:4/16:0represent the most common PC fatty acids (171 ).Consistent with the above data, particle content ofcholesteryl ester, free cholesterol, and phospholipidsubclasses including PC, phosphatidylethanolamine,phosphatidylinositol, sphingomyelin (SM), and lysoPCprogressively decreases with increase in hydrated den-sity from HDL2b to HDL3c (171 ). However, no suchdifferences are evident between HDL subspecies whendata for cholesteryl ester, PC, phosphatidylethanol-amine, phosphatidylinositol, and lysoPC are expressedas a percentage of total lipids, suggesting that their mo-lecular species are in dynamic equilibrium betweenHDL subpopulations. In a similar fashion, when lipid

HDL Nomenclature Special Report

Clinical Chemistry 57:3 (2011) 403

moieties of HDL are analyzed on the basis of theirtotal fatty acid content, the percentage distributions ofsaturated, monounsaturated, and polyunsaturated n-6and n-3 fatty acids are indistinguishable between HDLparticle subspecies (171 ).

However, the proportion of SM relative to totallipids decreases progressively in parallel with HDL den-sity from 12.8% in HDL2b to 6.2% in HDL3c. Conse-quently, the SM/PC molar ratio decreases from 0.38 inHDL2b to 0.18 in HDL3c. The distinctly low SM con-tent in HDL3c suggests that this pool is not in equilib-rium with that of other HDL subpopulations, consis-tent with the slow rate of exchange of SM betweenlipoproteins and cell membranes (172 ). The lowSM/PC ratio may reflect a distinct cellular origin(s) ofsmall HDL as suggested by the low SM content of smallnascent HDL particles secreted by J774 macrophages,which originate from the exofacial leaflet of the plasmamembrane (173 ).

Similarly to SM, free cholesterol content decreasestwofold from HDL2b to HDL3c (171, 174 ). As a result,the cholesteryl ester/free cholesterol ratio significantlyincreases with HDL density, supporting the contentionthat small HDL constitutes a major site of cholesterolesterification within the HDL particle spectrum (175 ).Increased LCAT activity and diminished SM/PC ratioin HDL3c are consistent with this proposal, becauseSM functions as a physiological inhibitor of LCAT(176, 177 ).

Among the minor bioactive lipid components, theabundance of sphingosine-1-phosphate (S1P) perHDL particle is asymmetric across the HDL spectrum,with preferential enrichment in HDL3 (40 –50 mmol/mol HDL) compared to HDL2 subfractions (15–20mmol/mol) (171, 177, 178 ). Enrichment of smallHDL3 in S1P might be mechanistically related to thepotent capacity of such particles to acquire polar lipidsof cellular origin (171 ).

The heterogeneity of the HDL lipidome maytranslate into distinct functionality of HDL sub-populations. Indeed, small, dense HDL3 exhibitmore potent antioxidative activity and antiinflam-matory activities compared to large, light HDL2 in-dependently of the concentration basis employed forsuch comparison (total protein, total mass, or parti-cle number) (179 ). Furthermore, small, dense HDL3exhibit more potent capacity to protect human mi-crovascular endothelial cells from apoptosis inducedby oxidized LDL compared to large, light HDL2 ir-respective of comparison method (177, 178 ). Fi-nally, small HDL particles represent more avidacceptors of cellular cholesterol via the ABCA1-dependent pathway (173, 175 ).

Studies of the mechanistic aspects of the potentantioxidative activity of HDL3 particles have revealed

that their capacity to inactivate LDL-derived lipid hy-droperoxides critically depends on the surface lipid flu-idity that is primarily determined by the HDL lipi-dome, and notably by the SM/PC ratio (179 ). Theenhanced fluidity of the surface monolayer of smallHDL3 particles related to the low abundance of SMmay equally contribute to their increased cellular cho-lesterol efflux capacity. Finally, the potent capacity ofHDL3 to protect endothelial cells from apoptosis mayin part reflect HDL3 enrichment in S1P, a minor bio-active lipid (171, 179 ).

Available data thereby suggest that lipidomic anal-yses of HDL particles can be used to obtain informa-tion regarding antiatherogenic functionality of HDL.Data are presently available on the relationship be-tween the HDL lipidome and cardiovascular risk, andthis information may prove useful for the evaluation ofnew HDL-increasing agents.

Summary

A growing body of evidence from epidemiological data,studies in animal models, and available clinical trialdata supports the contention that HDL represents thenext therapeutic target for reduction in the residual risktypical of statin-treated, high-risk cardiovascular pa-tients. Measurement of HDL cholesterol has been em-ployed as the principal method to assess the role ofHDL as a risk factor for CVD. The physicochemicaland functional heterogeneity of HDL presents an im-portant challenge to the cardiovascular field for devel-opment of more effective laboratory and clinicalmethods to quantify HDL with a predictive value inassessing cardiovascular risk. Moreover, the metabolicand clinical associations between low concentrations ofHDL cholesterol and increased concentrations ofcholesterol-depleted HDL particles and cholesterol-enriched triglyceride remnants must be considered inCVD risk assessment.

The initial gold standard for the separation ofplasma HDL was analytical ultracentrifugation withseparation of HDL initially into HDL2 and HDL3 withfurther resolution into HDL2a, HDL2b, and HDL3.The identification of the density profile of HDL pro-vided the information essential for the development ofpreparative methods to isolate, subfractionate, andcharacterize HDL. Concomitantly, characterization ofHDL particles by size was achieved with gradient gelelectrophoresis, which separates HDL into HDL2b,HDL2a, HDL3a, HDL3b, and HDL3c.

Further resolution of HDL particles by 2-D gelelectrophoresis into pre-� HDL and �1–�4 HDL hasbeen extremely useful in the characterization of HDLfrom animal models, clinical trials using differentdrugs, and genetic defects in lipoprotein metabolism.

Special Report

404 Clinical Chemistry 57:3 (2011)

The observation of the metabolic maturation of pre-�HDL into HDL �1 to �4, the identification of the pre-dictive value of reduction in �1 and cardiovascularrisk, and the HDL profile in the genetic dyslipopro-teinemias provided important new insights into HDLstructure and metabolism.

A major advance in the assessment of HDL hasbeen the development of methods to quantify HDLparticle number. The new ion mobility method toquantify plasma apo B lipoproteins as well as HDL israpidly progressing and will be useful in the clinicalassessment of the plasma lipoproteins. NMR holdsgreat promise for the quantification of the number ofHDL particles in clinical samples, analogous to the par-ticle number quantification of apo B– containing li-poproteins by using apo B immunoassay or NMR. Theability to correlate HDL particle number with HDLcholesterol, as well as the possibility of eliminating theconfounding influence of the apo B– containing li-poproteins in the assessment of the potential associa-tion with clinical events, will provide new insights intothe role of HDL in cardiovascular disease. Recent datafrom the VA-HIT and Multi-Ethnic Study of Athero-sclerosis cIMT (carotid intima-media thickness) clini-cal trials substantiate the potential importance of thisnew approach to HDL particle quantification andCVD risk.

The development of new mass spectrometricmethods has provided the unique opportunity to de-termine the protein composition of HDL and itsconstituent subfractions. In addition to the classicapolipoprotein, HDL contains proteins associatedwith inflammation, clotting, and complement regu-lation as well as proteolytic enzymes. Of particularinterest was the finding that there were clusters ofproteins on separate HDL particles, which gave riseto the hypothesis that specific subsets of HDL parti-cles may exert specific function(s). In this regard, theunique proteome of HDL3c is particularly effectivein protecting LDL against oxidation. The lipid com-ponents of HDL also exhibit marked heterogeneity.Thus the ratio of cholesteryl ester/free cholesteroland PC/SM differs among HDL subfractions, andthe ratio of PC/SM in the smallest HDL3c particlesmarkedly influences LCAT activation, the surface ri-gidity of the HDL particle, and, potentially, proteincomposition. In addition, bioactive lipid compo-nents such as S1P are preferentially associated withthe HDL3c particle subset.

The combined findings from multiple analyticalprocedures used to characterize HDL support the con-cept that the marked physicochemical heterogeneity ofHDL particles is the underlying basis for their func-tional heterogeneity. Further structural and composi-tional analysis of HDL particles may provide additional

information on the identification of HDL particleswith specific unique functions. Equally, studies at themolecular level possess the potential not only to revealnew risk biomarkers, but also to identify new pharma-cotherapeutic targets to reduce atherosclerosis andCVD.

To facilitate the future characterization of HDLsubfractions, development of a new uniform no-menclature for the subfractions of HDL is critical,and is proposed in this manuscript (Table 3). Thisclassification system defines 5 HDL subclasses on thebasis of physical and chemical properties, and as-signs very large HDL particles to the largest subclass,and large HDL, medium HDL, small HDL, and verysmall HDL to the smallest and most dense sub-classes. The very small HDL subclass includes pre-�,discoidal, or nascent HDL. The nomenclature will betested through analysis of split samples by using thevarious methods described in this report. We antic-ipate that the development of a uniform nomencla-ture for HDL subfractions will increase our ability tocompare data obtained with different methodologicapproaches for HDL fractionation, and to assess theclinical effects of different agents, which modulateHDL particle structure, metabolism, and function,and thus CVD risk. Prospective studies will be essen-tial in establishing the associations between HDLsubclasses and CVD that will be revealed by usingthese various methodologies (180 ).

Author Contributions: All authors confirmed they have contributed tothe intellectual content of this paper and have met the following 3 re-quirements: (a) significant contributions to the conception and design,acquisition of data, or analysis and interpretation of data; (b) draftingor revising the article for intellectual content; and (c) final approval ofthe published article.

Authors’ Disclosures or Potential Conflicts of Interest: Uponmanuscript submission, all authors completed the Disclosures of Poten-tial Conflict of Interest form. Potential conflicts of interest:

Employment or Leadership: M.J. Chapman, European Atheroscle-rosis Society; M.M. Hussain, Chylos; J.D. Otvos, LipoScience; E.J.Schaefer, Boston Heart Laboratory.Consultant or Advisory Role: R.S. Rosenson, Abbott Labs, Anthera,LipoScience, Residual Risk Reduction Initiative, Grain Foods Board,and Roche Genentech; H.B. Brewer, Jr., Merck, Merck-Schering-Plough, Schering-Plough, Roche, AstraZeneca, and Lilly; M.J. Chap-man, Merck, Pfizer; S. Fazio, Merck; R.M. Krauss, Merck, Roche,Metabolex, Corcept Pharmaceuticals, Celera, and Gilead; E.J. Schaefer,AstraZeneca, Arisaph, DuPont, Merck, Unilever, and Vatera.Stock Ownership: R.S. Rosenson, LipoScience; J.D. Otvos, Lipo-Science; E.J. Schaefer, Boston Heart Laboratory.Honoraria: R.S. Rosenson, Abbott Labs, Anthera Pharmaceuticals,LipoScience, Residual Risk Reduction Initiative, Roche-Genentech,XOMA, and Grain Foods Board; H.B. Brewer, Jr., Merck, Merck-Schering-Plough, Schering-Plough, Roche, AstraZeneca, and Lilly;S. Fazio, Merck; M.M. Hussain, Merck, GlaxoSmithKline, andPfizer; A. Kontush, Novo Nordisk; R.M. Krauss, Merck, Roche,Metabolex, Corcept Pharmaceuticals, Celera, and Gilead; E.J.

HDL Nomenclature Special Report

Clinical Chemistry 57:3 (2011) 405

Schaefer, AstraZeneca, Arisaph, DuPont, Merck, Unilever, andVatera.Research Funding: M.J. Chapman, Merck; S. Fazio, ISIS Gen-zyme; M.M. Hussain, NIH; A. Kontush, Pfizer, AstraZeneca, andGlaxoSmithKline; E.J. Schaefer, DuPont.Expert Testimony: None declared.

Other: R.M. Krauss, coinventor of 2 patents for lipoprotein subfrac-

tion analysis.

Role of Sponsor: The funding organizations played no role in the

design of study, choice of enrolled patients, review and interpretation

of data, or preparation or approval of manuscript.

References

1. Expert Panel on Detection, Evaluation, andTreatment of High Blood Cholesterol in Adults.Executive Summary of The Third Report of TheNational Cholesterol Education Program (NCEP)Expert Panel on Detection, Evaluation, andTreatment of High Blood Cholesterol In Adults(Adult Treatment Panel III). JAMA 2001;285:2486–97.

2. European guidelines on cardiovascular diseaseprevention in clinical practice: executive sum-mary. Eur J Cardiovasc Preven Rehab 2007;14(Suppl 2):E1–40.

3. Gotto AM, Brinton EA. Assessing low levels ofhigh-density lipoprotein cholesterol as a riskfactor in coronary heart disease: a workinggroup report and update. J Am Coll Cardiol2004;43:717–24.

4. Fruchart JC, Sacks F, Hermans MP, Assmann G.Brown V, Chapman J, et al. Executive statement,The Residual Risk Reduction Initiative: A call toaction to reduce residual vascular risk in dyslip-idemic patients. Diab Vasc Dis Res 2008;5:319–35.

5. Assmann G, Schulte H, von Eckardstein A et al.High-density lipoprotein cholesterol as a predic-tor of coronary heart disease risk. The PROCAMexperience and pathophysiological implicationsfor reverse cholesterol transport. Atherosclerosis1996;124(Suppl):S11–20.

6. Castelli WP. Cholesterol and lipids in the risk ofcoronary artery disease: the Framingham HeartStudy. Can J Cardiol 1988;(4 Suppl A):5A–10A.

7. Barter P, Gotto AM, LaRosa JC, Maroni J, SzarekM, Grundy SM, et al. HDL cholesterol, very lowlevels of LDL cholesterol, and cardiovascularevents. N Engl J Med 2007;357:1301–10.

8. Otvos JD, Jeyarajah EJ, Cromwell WC. Measure-ment issues related to lipoprotein heterogeneity.Am J Cardiol. 2002;90:22i–9i.

9. Brewer HB. HDL metabolism and the role of HDLin the treatment of high-risk patients with car-diovascular disease. Curr Cardiol Rep 2007;9:486–92.

10. Rosenson RS. Functional assessment of HDL:moving beyond static measures for risk assess-ment. Cardiovasc Drugs Ther 2010;24:71–5.

11. Santos RD, Asztalos BF, et al. Clinical presenta-tion, laboratory values, and coronary heart dis-ease risk in marked high-density lipoprotein-deficiency states. J Clin Lipidol 2008;2:237–47.

12. Kontush A, Chapman MJ. Functionally defectivehigh-density lipoprotein: a new therapeutic tar-get at the crossroads of dyslipidemia, inflamma-tion, and atherosclerosis. Pharmacol Rev 2006;58:342–74.

13. Asztalos BF, Sloop CH, Wong L, et al. Two-dimensional electrophoresis of plasmalipoproteins: recognition of new apo A-I-containing subpopulations. Biochim BiophysActa 1993;1169:291–300.

14. Rothblatt GH, de lL-M, Atger V, et al. Cellcholesterol efflux: integration of old and newobservations provides new insights. J Lipid Res1999;40:781–96.

15. Friedewald WT, Levy RI, Fredrickson DS. Estima-tion of the concentration of low-density lipopro-tein cholesterol in plasma, without use of thepreparative ultracentrifuge. Clin Chem 1972;18:499–502.

16. Emerging Risk Factors Collaboration, Di An-gelantonio E, Sarwar N, Perry P, Kaptoge S, RayKK, et al. Major lipids, apolipoproteins and riskof vascular disease. JAMA 2009;302:1993–2000.

17. Zhang L, Qiao Q, Tuomilehto J, Hammar N,Ruotolo G, Stehouwer CD, et al. DECODE StudyGroup. The impact of dyslipidaemia on cardio-vascular mortality in individuals without a priorhistory of diabetes in the DECODE Study. Ath-erosclerosis 2009;206:298–302.

18. Baigent C, Keech A, Kearney PM, Blackwell L,Buck G, Pollicino C, et al. Cholesterol TreatmentTrialists’ (CTT) Collaborators. Efficacy and safetyof cholesterol-lowering treatment: prospectivemetaanalysis of data from 90,056 participantsin 14 randomised trials of statins. Lancet 2005;366:1267–78.

19. Ridker PM, Genest J Boekholdt SM, Libby P,Gotto AM, Nordestgaard BG, et al. HDL choles-terol and residual risk of first cardiovascularevents after treatment with potent statintherapy: an analysis from the JUPITER trial. Lan-cet 2010;376:333–9.

20. Briel M, Ferreira-Gonzalez I, You JJ, KaranicolasPJ, Akl EA, Wu P, et al. Association betweenchange in high density lipoprotein cholesteroland cardiovascular disease morbidity andmortality: systematic review and meta-regression analysis. BMJ 2009;338:b92.

21. Miller WG, Myers GL, Sakurabayashi I, BachmanLM, Caudill SP, Dziekonski A, et al. Seven directmethods for measuring HDL and LDL cholesterolcompared with ultracentrifugation referencemeasurement procedures. Clin Chem 2010;56:977–86.

22. Havel RJ, Eder HA, Bragdon JH. The distributionand chemical composition of ultracentrifugallyseparated lipoproteins in human serum. J ClinInvest 1955;34:1345–53.

23. Remaley AT, Warnick GR. High densitylipoprotein: what is the best way to measure itsanti-atherogenic function? Exp Opin Med Diagn2008;2:773–88.

24. Warnick GR, Nauck M, Rifai N. Evolution ofmethods for measurment of HDL-cholesterol:from ultracentrifugation to homogenous assays.Clin Chem 2001;47:1579–96.

25. Langlois MR, Blaton VH. Historical milestones inmeasurment of HDL-cholesterol: impact on clin-ical and laboratory practice. Clin Chim Acta

2006;369:168–78.26. Gofman JW, Young W, Tandy R. Ischemic heart

disease, atherosclerosis, and longevity. Circula-tion 1966;34:679–97.

27. Anderson DW, Nichols AV, Pan SS, Lindgren FT.High density lipoprotein distribution. Resolutionand determination of three major componentsin a normal population sample. Atherosclerosis1978;29:161–79.

28. Williams PT. Feldman DE. Prospective study ofcoronary heart disease vs HDL2, HDL3, andother lipoproteins in Gofman’s Livermore Co-hort. Atherosclerosis 2011;214:196–202.

29. Blanche PJ, Gong EL, Forte TM, Nichols AV.Characterization of human high-density lipopro-teins by gradient gel electrophoresis. BiochimBiophys Acta 1981;665:408–19.

30. Johansson J, Carlson LA, Landou C, Hamsten A.High density lipoproteins and coronary athero-sclerosis. A strong inverse relation with the larg-est particles is confined to normotriglyceridemicpatients. Arterioscler Thromb 1991;11:174–82.

31. Berneis KK, Krauss RM. Metabolic origins andclinical significance of LDL heterogeneity. J LipidRes 2002;43:1363–79.

32. Asztalos BF, Cupples LA, Demissie S, HorvathKV, Cox CE, Batista MC, Schaefer EJ. High-density lipoprotein subpopulation profile andcoronary heart disease prevalence in male par-ticipants in the Framingham Offspring Study.Arterioscler Thromb Vasc Biol 2004;24:2181–7.

33. Chapman MJ, Goldstein S, Lagrange D, LaplaudPM. A density gradient ultracentrifugal proce-dure for the isolation of the major lipoproteinclasses from human serum. J Lipid Res 1981;22:339–58.

34. Goulinet S, Chapman MJ. Plasma LDL and HDLsubspecies are heterogenous in particle contentof tocopherols and oxygenated and hydrocar-bon carotenoids. Relevance to oxidative resis-tance and atherogenesis. Arterioscler ThrombVasc Biol 1997;17:786–96.

35. Guerin M, Bruckert E, Dolphin PJ, Turpin G,Chapman MJ. Fenofibrate reduces plasma cho-lesteryl ester transfer from HDL to VLDL andnormalizes the atherogenic, dense LDL profile incombined hyperlipidemia. Arterioscler ThrombVasc Biol 1996;16:763–72.

36. Kontush A, Chantepie S, Chapman MJ. Small,dense HDL particles exert potent protection ofatherogenic LDL against oxidative stress. Arte-rioscler Thromb Vasc Biol 2003;23:1881–8.

37. Cone JT, Segrest JP, Chung BH, Ragland JB,Sabesin SM, Glasscock A. Computerized rapidhigh resolution quantitative analysis of plasmalipoproteins based upon single vertical spin cen-trifugation. J Lipid Res 1982;23:923–35.

38. Kulkarni KR, Marcovina SM, Krauss RM, GarberDW, Glasscock AM, Segrest JP. Quantification ofHDL2 and HDL3 cholesterol by the Vertical Auto

Special Report

406 Clinical Chemistry 57:3 (2011)

Profile-II (VAP-II) methodology. J Lipid Res1997;38:2353–64.

39. Kulkarni KR. Cholesterol profile measurement byvertical auto profile method. Clin Lab Med 2006;26:787–802.

40. Ensign W, Hill N, Heward CB. Disparate LDLphenotypic classification among 4 differentmethods assessing LDL particle characteristics.Clin Chem 2006;52:1722–7.

41. Asztalos BF, Schaefer EJ, Horvath KV, YamashitaS, Miller M, Franceschini G, Calabresi L. Role ofLCAT in HDL remodeling: an investigation inLCAT deficiency states. J Lipid Res 2007;48:592–9.

42. Asztalos BF, De la Llera-Moya M, Dallal GE,Horvath KV, Schaefer EJ, Rothblat GH. Differen-tial effects of HDL subpopulations on cellularABCA1 and SRB1-mediated cholesterol efflux. JLipid Res 2005;46:2246–53.

43. Sankaranarayanan S, Oram JF, Asztalos BF,Vaughan AM, Lund-Katz S, Adorni MP, et al.Effect of acceptor composition and mechanismsof ABCG1-mediated cellular free cholesterol ef-flux. J Lipid Res 2009;50:275–84.

44. De la Llera-Moya M, Drazul-Schrader D, Aszta-los BF, Cuchel M, Rader DJ, Rothblat GH. Theability to promote efflux via ABCA1 determinesthe capacity of serum specimens with similarhigh density lipoprotein cholesterol to removecholesterol from macrophages. ArteriosclerThromb Vasc Biol 2010;30:796–801.

45. Sacks FM, Rudel LL, Conner A, Akeefe HJ, Ko-stner G, Baki T, et al. Selective delipidation ofplasma HDL enhances reverse cholesterol trans-port in vivo. J Lipid Res, 2009;50:894–907.

46. Santos RD, Schaefer EJ, Asztalos BF, Polisecki E,Hegele RA, Martinez L, et al. Characterization ofhigh density lipoprotein particles in familial apoA-I deficiency with premature coronary athero-sclerosis, corneal arcus and opacification, andtubo-eruptive and planar xanthomas. J Lipid Res2008;49:349–57.

47. Asztalos BF, Brousseau ME, McNamara JR, Hor-vath KV, Roheim PS, Schaefer EJ. Subpopula-tions of high-density lipoproteins in homozy-gous and heterozygous Tangier disease.Atherosclerosis 2001;156:217–25.

48. Lamon-Fava S, Asztalos BF, Howard TD, Re-boussin DM, Horvath KV, Schaefer EJ, Her-rington DM. Association of polymorphisms ingenes involved in lipoprotein metabolism withplasma concentrations of remnant lipoproteinsand HDL subpopulations before and after hor-mone therapy in postmenopausal women. ClinEndocrinol (Oxf) 2010;72:169–75.

49. Asztalos BF, Horvath KV, Kajinami K, NartsuphaC, Cox CE, Batista M, et al. Apo composition ofHDL in cholesteryl ester transfer protein defi-ciency. J Lipid Res 2004;45:448–55.

50. Asztalos BF, Collins D, Cupples LA, Demissie S,Horvath KV, Bloomfield HE, et al. Value of highdensity lipoprotein (HDL) subpopulations in pre-dicting recurrent cardiovascular events in theVeterans Affairs HDL Intervention Trial. Arterio-scler Thromb Vasc Biol 2005;25:2185–91.

51. Asztalos BF, Batista M, Horvath KV, Cox CE,Dallal GE, Morse JS, et al. Change in alpha 1HDL concentration predicts progression in coro-nary artery stenosis. Arterioscler Thromb VascBiol 2003;23:847–52.

52. Brousseau ME, Diffenderfer MR, Millar JS, Nart-supha C, Asztalos BF, Welty FK, et al. Effects ofcholesteryl ester transfer protein inhibition onhigh-density lipoprotein subspecies, apolipopro-tein A-I metabolism, and fecal sterol excretion.Arterioscler Thromb Vasc Biol 2005;25:1057–64.

53. Lamon-Fava S, Diffenderfer MR, Barrett PHR,Buchsbaum A, Nyaku M, Horvath K, et al.Extended-release niacin alters the metabolismof plasma apolipoprotein (apo) A-I and apoB-containing lipoproteins. Arterioscler ThrombVasc Biol 2008;28:1672–8.

54. Asztalos BF, Swarbrick MM, Schaefer EJ, DallalGE, Horvath KV, Ai M, et al. Effects of weightloss, induced by gastric bypass surgery, on HDLremodeling in obese women. J Lipid Res 2010;51:2405–12.

55. Asztalos BF, LeMaulf F, Dallal GE, Stein E, JonesPH, Horvath KV, et al. Comparison of the effectsof high doses of rosuvastatin versus atorvastatinon the subpopulations of high density lipopro-teins. Am J Cardiol 2007;99:681–5.

56. Lamon-Fava S. Diffenderfer MR, Barrett PH,Buchsbaum A, Matthan NR, Lichtenstein AH,et al. Effects of different doses of atorvastatinon human apolipoprotein B-100, B-48, and A-Imetabolism. J Lipid Res 2007;48:1746–53.

57. Asztalos BF, Horvath KV, McNamara JR, RoheimPS, Rubenstein JJ, Schaefer EJ. Comparing theeffects of five different statins on the HDL sub-population profiles of coronary heart diseasepatients. Atherosclerosis 2002;164:361–9.

58. Lounila J, Ala-Korpela M, Jokisaari J. Effects oforientational order and particle size on the NMRline positions of lipoproteins. Phys Rev 1994;72:4049–52.

59. Otvos JD. Measurement of lipoprotein subclassprofiles by nuclear magnetic resonance spec-troscopy. Clin Lab 2002;48:171–80.

60. Jeyarajah EJ, Cromwell WC, Otvos JD. Lipopro-tein particle analysis by nuclear magnetic reso-nance spectroscopy. Clin Lab Med 2006;26:847–70.

61. Freedman DS, Otvos JD, Jeyarajah EJ, Shalau-rova I, Cupples LA, Parise H, et al. Sex and agedifferences in lipoprotein subclasses measuredby nuclear magnetic resonance spectroscopy:The Framingham Study. Clin Chem 2004;50:1189–200.

62. Mora S, Szklo M, Otvos JD, Greenland P, PsatyBM, Goff DC Jr, et al. LDL particle subclasses,LDL particle size, and carotid atherosclerosis inthe Multi-Ethnic Study of Atherosclerosis. Ath-erosclerosis 2007;192:211–7.

63. Barzilai N, Atzmon G, Schechter C, Schaefer EJ,Cupples AL, Lipton R, et al. Unique lipoproteinphenotype and genotype associated with excep-tional longevity. JAMA 2003;290:2030–40.

64. Garvey WT, Kwon S, Zheng D, Shaughnessy S,Wallace P, Hutto A, et al. The effects of insulinresistance and Type 2 diabetes mellitus on li-poprotein subclass particle size and concentra-tion determined by nuclear magnetic resonance.Diabetes 2003;52:453–62.

65. Goff DC, D’Agostino RB Jr., Haffner SM, OtvosJD. Insulin resistance and adiposity influencelipoprotein size and subclass concentrations: re-sults from the Insulin Resistance AtherosclerosisStudy. Metabolism 2005;54:264–70.

66. Festa A, Williams K, Hanley AJG, Otvos JD, GoffDC, Wagenknecht LE, Haffner SM. Nuclear mag-netic resonance lipoprotein abnormalities in pre-diabetic subjects in the Insulin Resistance Ath-erosclerosis Study (IRAS). Circulation 2005;111:3465–72.

67. Kathiresan S, Otvos JD, Sullivan LM, Keyes MJ,Schaefer EJ, Wilson PW, et al. Increased smalllow-density lipoprotein particle number: aprominent feature of the metabolic syndrome inthe Framingham Heart Study. Circulation 2006;113:20–9.

68. Rosenson RS, Freedman DS, Otvos JD. Relationsof lipoprotein subclass levels and LDL size toprogression of coronary artery disease in thePLAC I trial. Am J Cardiol 2002;90:89–94.

69. Kuller L, Arnold A, Tracy R, Otvos J, Burke G,Psaty B, et al. NMR spectroscopy of lipoproteinsand risk of CHD in the Cardiovascular HealthStudy. Arterioscler Thromb Vasc Biol 2002;22:1175–80.

70. Otvos JD, Collins D, Freedman DS, Shalaurova I,Schaefer EJ, McNamara JR, et al. LDL and HDLparticle subclasses predict coronary events andare changed favorably by gemfibrozil therapy inthe Veterans Affairs HDL Intervention Trial (VA-HIT). Circulation 2006;113:1556–63.

71. Birjmohun RS, Dallinga-Thie GM, KuivenhovenJA, Stroes ES, Otvos JD, Wareham NJ, et al. ApoA-II is inversely associated with risk of futurecoronary artery disease. Circulation 2007;116:2029–35.

72. van der Steeg WA, Holme I, Boekholdt SM,Larsen ML, Lindahl C, Stroes ESG, et al. High-density lipoprotein cholesterol, high-density li-poprotein particle size, and apo A-1: Signifi-cance for cardiovascular risk. J Am Coll Cardiol2008;51:634–42.

73. Mora S, Otvos JD, Rifai N, Rosenson RS, BuringJE, Ridker PM. Lipoprotein particle profiles bynuclear magnetic resonance compared withstandard lipids and apos in predicting cardio-vascular disease in women. Circulation 2009;119:931–9.

74. El Harchaoui K, Arsenault BJ, Franssen R, De-spres J-P, Hovingh GK, Stroes ESG, et al. High-density lipoprotein particle size and concentra-tion and coronary risk. Ann Intern Med 2009;150:84–93.

75. Kraus WE, Houmard JA, Duscha BD, KnetzgerKJ, Wharton MB, McCartney JS, et al. Effects ofthe amount and intensity of exercise training onplasma lipoproteins. N Engl J Med 2002;347:1483–92.

76. Schaefer EJ, McNamara JR, Taylor T, Daly JA,Gleason JA, Seman LJ, et al. Effects of atorva-statin on fasting and postprandial lipoproteinsubclasses in coronary heart disease patientsversus control subjects. Am J Cardiol 2002;90:689–96.

77. Morgan JM, Capuzzi DM, Baksh RI, Intenzo C,Carey CM, Reese D, Walker K. Effects ofextended-release niacin on lipoprotein subclassdistribution. Am J Cardiol 2003;91:1432–6.

78. Ikewaki K, Tohyama J, Nakata Y, Wakikawa T,Kido T, Mochizuki S. Fenofibrate effectively re-duces remnants, and small dense LDL, and in-creases HDL particle number in hypertriglyceri-demic men: a nuclear magnetic resonancestudy. J Atheroscler Thromb 2004;11:278–85.

HDL Nomenclature Special Report

Clinical Chemistry 57:3 (2011) 407

79. Szapary PO, Bloedon LT, Samaha FH, Duffy D,Wolfe ML, Soffer D, et al. Effects of pioglitazoneon lipoproteins, inflammatory markers, and adi-pokines in nondiabetic patients with metabolicsyndrome. Arterioscler Thromb Vasc Biol 2006;26:182–8.

80. Kuvin JT, Dave DM, Sliney KA, Mooney P, PatelAR, Kimmelsteil CD, Karas RH. Effects ofextended-release niacin on lipoprotein particlesize, distribution, and inflammatory markers inpatients with coronary artery disease. Am J Car-diol 2006;98:743–5.

81. Davidson MH, McKenney JM, Shear CL, RevkinJH. Efficacy and safety of torcetrapib, a novelcholesteryl ester transfer protein inhibitor, inindividuals with below-average high-density li-poprotein cholesterol levels. J Am Coll Cardiol2006;48:1774–81.

82. McKenney JH, Davidson MH, Shear CL, RevkinJH. Efficacy and safety of torcetrapib, a novelcholesteryl ester transfer protein inhibitor, inindividuals with below-average high-density li-poprotein cholesterol levels on a background ofatorvastatin. J Am Coll Cardiol 2006;48:1782–90.

83. Deeg MA, Buse JB, Goldberg RB, Kendall DM,Zagar AJ, Jacober SJ, et al. Pioglitazone androsiglitazone have different effects on serumlipoprotein particle concentrations and sizes inpatients with type 2 diabetes and dyslipidemia.Diabetes Care 2007;30:2458–64.

84. Rosenson RS, Otvos JD, Hsia J. Effects of rosu-vastatin and atorvastatin on low-density andhigh-density lipoprotein particle concentrationsin patients with metabolic syndrome: a random-ized, double-blind, controlled study. DiabetesCare 2009;32:1087–91.

85. Caulfield MP, Li S, Lee G, Blanche PJ, SalamehWA, Benner WH, et al. Direct determination oflipoprotein particle sizes and concentrations byion mobility analysis. Clin Chem 2008;54:1307–16.

86. Musunuru K, Orho-Melander M, Caulfield MP, LiS, Salameh WA, Reitz RE, et al. Ion mobilityanalysis of lipoprotein subfractions identifiesthree independent axes of cardiovascular risk.Arterioscler Thromb Vasc Biol 2009;29:1975–80.

87. Contois J, McNamara JR, Lammi-Keefe C, Wil-son PW, Massov T, Schaefer EJ. Reference inter-vals for plasma apo A-1 determined with astandardized commercial immunoturbidimetricassay: results from the Framingham OffspringStudy. Clin Chem 1996;42:507–14.

88. Eggerman TL, Hoeg JM, Meng MS, Tombragel A,Bojanovski D, Brewer HB Jr. Differential tissue-specific expression of human apoA-I and apoA-II. J Lipid Res 1991;32:821–8.

89. Walldius G, Jungner I, Holme I, Aastveit AH,Kolar W, Steiner E. High apo B, low apo A-I, andimprovement in the prediction of fatal myocar-dial infarction (AMORIS study): a prospectivestudy. Lancet 2001;358:2026–33.

90. Walldius G, Jungner I. Apo A-I versus HDL cho-lesterol in the prediction of risk for myocardialinfarction and stroke. Curr Opin Cardiol 2007;22:359–67.

91. Mora S, Cui Y, Ridker PM. Association of HDLcholesterol and apolipoprotein A-1 with incidentCVD across levels of LDL cholesterol in 27,748

women. Circulation. 2009;120:S507–8. AHAAbstract 1462.

92. Gotto AM, Jr., Whitney E, Stein EA, Shapiro DR,Clearfield M, Weis S, et al. Relation betweenbaseline and on-treatment lipid parameters andfirst acute major coronary events in the AirForce/Texas Coronary Atherosclerosis Preven-tion Study (AFCAPS/TexCAPS). Circulation 2000;101:477–84.

93. Marcovina SM, Albers JJ, Henderson LO, Han-non WH. International Federation of ClinicalChemistry standardization project for measure-ments of apos A-I and B. III: comparability ofapo A-I values by use of international referencematerial. Clin Chem 1993;39:773–81.

94. Ferretti G, Bacchetti T, Negre-Salvayre A, Sal-vayre R, Dousset N, Curatola G. Structural mod-ifications of HDL and functional consequences.Atherosclerosis 2006;184:1–7.

95. Shao B, Oda MN, Vaisar T, Oram JF, HeineckeJW. Pathways for oxidation of high-density li-poprotein in human cardiovascular disease. CurrOpin Mol Ther 2006;8:198–205.

96. Zheng L, Nukuna B, Brennan ML, Sun M, Goor-mastic M, Settle M, et al. Apo A-I is a selectivetarget for myeloperoxidase-catalyzed oxidationand functional impairment in subjects with car-diovascular disease. J Clin Invest 2004;114:529–41.

97. Alaupovic P, Lee DM, McConathy WJ. Studieson the composition and structure of plasmalipoproteins: distribution of lipoprotein familiesin major density classes of normal humanplasma lipoproteins. Biochim Biophys Acta1972;260:689–707.

98. Alaupovic P. The concept of apolipoprotein-defined lipoprotein families and its clinical sig-nificance. Curr Atheroscler Rep 2003;5:459–67.

99. James RW, Hochstrasser D, Tissot JD, Funk M,Appel R, Barja F, et al. Protein heterogeneity oflipoprotein particles containing apoA-I withoutapoA-II and apoA-I with apoA-II isolated fromhuman plasma. J Lipid Res. 1988;29:1557–71.

100. Castro GR, Fielding CJ. Early incorporation ofcell-derived cholesterol into pre-�-migratinghigh-density lipoprotein. Biochemistry 1988;27:25–9.

101. Miida T, Kawano M, Fielding CJ, Fielding PE.Regulation of the concentration of pre-betahigh-density lipoprotein in normal plasma bycell membranes and lecithin-cholesterol acyl-transferase activity. Biochemistry 1992;31:11112–7.

102. Weng W, Brandenburg NA, Zhong S, Halkias J,Wu L, Jiang XC, et al. ApoA-II maintains HDLlevels in part by inhibition of hepatic lipase:studies in apoA-II and hepatic lipase doubleknockout mice. J Lipid Res 1999;40:1064–70.

103. De Beer MC, Durbin DM, Cai, L, Jonas A, deBeer FC, van der Westhuyzen DR. Apolipopro-tein A-II modulates the binding and selectiveuptake of reconstituted HDL by scavenger re-ceptor BI. J Biol Chem 2001;276:15832–9.

104. Wang N, Lan D, Chen W, Matsuura F, Tall AR.ATP-binding cassette transporters G1 and G4mediate cellular cholesterol efflux to high-density lipoproteins. Proc Natl Acad Sci USA2004:31:9774–9.

105. Kennedy MA, Barrera GC, Nakamura K, BaldanA, Tarr P, Fishbein MC, et al. ABCG1 has a

critical role in mediating cholesterol efflux toHDL and preventing cellular lipid accumulation.Cell Metab 2005;1:121–31.

106. Matsuura F, Wang N, Chen W, Jiang XC, TallAR. HDL from CETP-deficient subjects show en-hanced ability to promote cholesterol effluxfrom macrophages in an apoE-and ABCG1-dependent pathway. Clin Invest 2006;116,1435–42.

107. Yancey PG, Bortnick AE, Kellner-Weibel G, de laLlera-Moya M, Phillips MC, Rothblat GH. Impor-tance of different pathways of cellular choles-terol efflux. Arterioscler Thomb Vasc Biol 2003;23:712–9.

108. Williams DL, Connelly MA, Temel RE, SwarnakarS, Phillips MC, de la Llera-Moya M, Rothblat GH.Scavenger receptor B1 and cholesterol traffick-ing. Curr Opin Lipidol 1999;10:329–39.

109. Fruchart JC, Ailhaud G. Apolipoprotein A-containing lipoprotein particles: physiologicalrole, quantification and clinical significance. ClinChem 1992;38:793–7.

110. Puchois P, Kandoussi A, Fievet P, Fourrier JL,Bertrand M, Koren E, Fruchart JC. Apolipopro-tein A-I containing lipoproteins in coronary ar-tery disease. Atherosclerosis 1987;68:35–40.

111. Genest JJ, Bard JM, Fruchart JC, Ordovas JM,Wilson PF, Schaefer EJ. Plasma apolipoproteinAI, A-II, B, E and C-III containing particles inmen with premature coronary artery disease.Atherosclerosis 1991;90:149–57.

112. Montali A, Vega GL, Grundy SM. Concentrationsof apolipoprotein A-I-containing particles in pa-tients with hypoalphalipoproteinemia. Arterio-scler Thromb 1994:14,511–7.

113. Parra HJ, Arveiler D, Evans AE, Cambou JP,Amouyel P, Bingham A, et al. A case-controlstudy of lipoprotein particles in two populationsat contrasting risk for coronary heart disease:the ECTIM study. Arterioscler Thromb 1992;12:701–7.

114. Luc G, Bard JM, Ferrires J, Evans A, Amouyel P,Arveiler D, et al. on behalf of the PRIME studyGroup. Value of HDL cholesterol, apolipoproteinA-I, lipoprotein A-I, and lipoprotein A-I/A-II inprediction of coronary heart disease: the PRIMEStudy. Arterioscler Thromb Vasc Biol 2002;22:1155–61.

115. Asztalos B, Zhang W, Roheim PS, Wong L. Roleof free apolipoprotein A-I in cholesterol efflux.Formation of pre-alpha-migrating high-densitylipoprotein particles. Arterioscler Thromb VascBiol 1997;17:1630–6.

116. Rader DJ, Castro G, Zech LA, Fruchart JC,Brewer HB Jr. In vivo metabolism of apolipopro-tein A-I on high density lipoprotein particlesLpA-I and LpA-I,A-II. J Lipid Res 1991;32:1849–59.

117. Zech LA, Schaefer EJ, Bronzert TJ, Aamodt RL,Brewer HB Jr. Metabolism of human apolipopro-teins A-I and A-II: compartmental models. JLipid Res 1983;24:60–71.

118. Ji J, Watts GF, Johnson AG. Chan DC, Ooi EM,Rye KA, et al. High-density lipoprotein (HDL)transport in the metabolic syndrome: Applica-tion of a new model for HDL particle kinetics.J Clin Endo Metab 2006;91:973–9.

119. deBeer MC, Webb NR, Whitaker NL, WroblewskiJM, Jahangiri A, et al. SR-BI selective lipiduptake: subsequent metabolism of acute phase

Special Report

408 Clinical Chemistry 57:3 (2011)

HDL. Arterioscler Thromb Vasc Biol 2009;29:1298–303.

120. Ng DS, Vezina C, Wolever TS, Hegele RA, Con-nelly PW. Apolipoprotein A-I deficiency.Biochemical and metabolic characteristics.Arterioscler Thromb Vasc Biol 1995;15:2157–64.

121. Ikewaki K, Matsunaga A, Han H, Watanabe H,Endo A, Tohyama J, et al. A novel two nucleo-tide deletion in the apolipoprotein A-I gene,apoA-I Shinbashi, associated with high densitylipoprotein deficiency, corneal opacities, planarxanthomas, and premature coronary artery dis-ease. Atherosclerosis 2004;172:39–45.

122. Dastani Z, Dangoisse C, Boucher B, Desbiens K,Krimbou L, Dufour R, et al. A novel nonsenseapolipoprotein AI mutation (apoA-I(E136X))causes low HDL cholesterol in French Canadi-ans. Atherosclerosis 2006;185:127–36.

123. Deeb SS, Takata K, Peng RL, Kajiyama G, AlbersJJ. A splice-junction mutation responsible forfamilial apolipoprotein A-II deficiency. Am JHum. Genet 1990;46:822–7.

124. Schaefer EJ, Blum CB, Levy RI, Jenkins LL,Alaupovic P, Foster DM, Brewer HB Jr. Metab-olism of high-density lipoprotein apolipopro-teins in Tangier disease. N Engl J Med 1978;299:905–10.

125. Rader DJ, Ikewaki K, Duverger N, Schmidt H,Pritchard H, Frohlich J, et al. Markedly acceler-ated catabolism of apolipoprotein A-II (Apo A-II)and high density lipoproteins containing ApoA-II in classic lecithin: cholesterol acyltrans-ferase deficiency and fish-eye disease. Clin In-vest 1994;93:321–30.

126. Calabresi L, Pisciotta L, Costantin A, Frigerio I,Eberini I, Alessandrini P, et al. The molecularbasis of lecithin:cholesterol acyltransferase de-ficiency syndromes: a comprehensive study ofmolecular and biochemical findings in 13 unre-lated Italian families. Arterioscler Thromb VascBiol 2005;9:1972–8.

127. Ikewaki K, Rader DJ, Sakamoto T, Nishiwaki M,Wakimoto N, Schaefer JR, et al. Delayed catab-olism of high density lipoprotein apolipoproteinsA-I and A-II in human cholesteryl ester transferprotein deficiency. J Clin Invest 1993;92:1650–8.

128. Ikewaki K, Nishiwaki M, Sakamoto T, IshikawaT, Fairwell T, Zech LA, et al. Increased catabolicrate of low density lipoproteins in humans withcholesteryl ester transfer protein deficiency.J Clin Invest 1995;96:1573–81.

129. Brinton EA, Eisenberg S, Breslow JL. Increasedapo A-I and apo A-II fractional catabolic rate inpatients with low high density lipoprotein-cholesterol levels with or without hypertriglyc-eridemia. J Clin Invest 1991;87:536–44.

130. Horowitz BS, Goldberg IJ, Merab J, Vanni TM,Ramakrishnan R, Ginsberg HN. Increasedplasma and renal clearance of an exchangeablepool of apolipoprotein A-I in subjects with lowlevels of high density lipoprotein cholesterol.J Clin Invest 1993;91:1743–52.

131. Chapman MJ, Le Goff W, Guerin M, Kontush A.Cholesteryl ester transfer protein: at the heart ofthe action of lipid-modulating therapy withstatins, fibrates, niacin, and cholesteryl estertransfer protein inhibitors Eur Heart J 2010;31:149–64.

132. Brinton EA, Eisenberg S, Breslow JL. Human HDLcholesterol levels are determined by apoA-I frac-tional catabolic rate which correlates inverselywith estimates of HDL particle size. Effects ofgender, hepatic and lipoprotein lipases, triglyc-eride, insulin levels and body fat distribution.Arterioscler Thromb Vasc Biol 1994;14:707–20.

133. Chan DC, Watts GF, Ooi EM, Rye KA, Ji J,Johnson AG, Barrett PH. Regulatory effects offenofibrate and atorvastatin on lipoprotein A-Iand lipoprotein A-I:A-II kinetics in the metabolicsyndrome. Diabetes Care 2009;32:2111–3.

134. Barter PJ, Brandrup-Wognsen G, Palmer MK,Nicholls SJ. Effect of statins on HDL: a complexprocess unrelated to changes in LDL: Analysis ofthe VOYAGER Database. J Lipid Res 2010;51:1546–53.

135. Vu-Dac N, Schoonjans K, Kosykh V, Dallon-geville J, Fruchart JC, Staels B, Auwerx J. Fi-brates increase human apolipoprotein A-II ex-pression through activation of the peroxisomeproliferator activated receptor. J Clin Invest.1995;96:741–50.

136. Bradley WA, Gianturco SH. ApoE is necessaryand sufficient for the binding of largetriglyceride-rich lipoproteins to the LDL receptor;apoB is unnecessary. J Lipid Res 1986;27:40–8.

137. Ehnholm C, Mahley RW, Chappell DA, Weisgra-ber KH, Ludwig E, Witztum JL. Role of apoli-poprotein E in the lipolytic conversion of beta-very low density lipoproteins to low densitylipoproteins in type III hyperlipoproteinemia.Proc Natl Acad Sci U S A 1984;81:5566–70.

138. Weisgraber KH. Apolipoprotein E distributionamong human plasma lipoproteins: role of thecysteine-arginine interchange at residue 112. JLipid Res 1990;31:1503–11.

139. Parks BW, Srivastava R, Yu S, Kabarowski JH.ApoE-dependent modulation of HDL and ath-erosclerosis by G2A in LDL receptor-deficientmice independent of bone marrow-derived cells.Arterioscler Thromb Vasc Biol 2009;29:539–47.

140. Funke H, Boyles J, Weisgraber KH, Ludwig EH,Hui DY, Mahley RW. Uptake of apolipoproteinE-containing high density lipoproteins by he-patic parenchymal cells. Arteriosclerosis 1984;4:452–61.

141. Mahley RW, Weisgraber KH, Innerarity T,Brewer HB Jr, Assmann G. Swine lipoproteinsand atherosclerosis. Changes in the plasma li-poproteins and apoproteins induced by choles-terol feeding. Biochemistry 1975;14:2817–23.

142. Hatters DM, Peters-Libeu CA, Weisgraber KH.Apolipoprotein E structure: insights into func-tion. Trends Biochem Sci 2006;31:445–54.

143. Yancey PG, Yu H, Linton MF, Fazio S. Apathway-dependent on apoE, ApoAI, andABCA1 determines formation of buoyant high-density lipoprotein by macrophage foam cells.Arterioscler Thromb Vasc Biol 2007;27:1123–31.

144. Su YR, Blakemore JL, Zhang Y, Linton MF, FazioS. Lentiviral transduction of apoAI into hemato-poietic progenitor cells and macrophages: appli-cations to cell therapy of atherosclerosis. Arte-rioscler Thromb Vasc Biol 2008;28:1439–46.

145. Ishiguro H, Yoshida H, Major AS, Zhu T, BabaevVR, Linton MF, Fazio S. Retrovirus-mediatedexpression of apolipoprotein A-I in the macro-phage protects against atherosclerosis in vivo.

J Biol Chem 2001;276:36742–8.146. Major AS, Dove DE, Ishiguro H, Su YR, Brown

AM, Liu L, et al. Increased cholesterol efflux inapolipoprotein AI (ApoAI)-producing macro-phages as a mechanism for reduced atheroscle-rosis in ApoAI((-/-)) mice. Arterioscler ThrombVasc Biol 2001;21:1790–5.

147. Mahley RW, Innerarity TL, Bersot TP, Lipson A,Margolis S. Alterations in human high-densitylipoproteins, with or without increased plasma-cholesterol, induced by diets high in cholesterol.Lancet 1978;2:807–9.

148. Neale MC, de Knijff P, Havekes LM, BoomsmaDI. ApoE polymorphism accounts for only part ofthe genetic variation in quantitative ApoE lev-els. Genet Epidemiol 2000;18:331–40.

149. Yamashita S, Sprecher DL, Sakai N, MatsuzawaY, Tarui S, Hui DY. Accumulation of apolipopro-tein E-rich high density lipoproteins in hyperal-phalipoproteinemic human subjects withplasma cholesteryl ester transfer protein defi-ciency. J Clin Invest 1990;86:688–95.

150. Matsuura F, Wang N, Chen W, Jiang XC, TallAR. HDL from CETP-deficient subjects showsenhanced ability to promote cholesterol effluxfrom macrophages in an apoE- and ABCG1-dependent pathway. J Clin Invest 2006;116:1435–42.

151. Tall AR. CETP inhibitors to increase HDL choles-terol levels. N Engl J Med 2007;356:1364–6.

152. Young EK, Chatterjee C, Sparks DL. HDL-ApoEcontent regulates the displacement of hepaticlipase from cell surface proteoglycans. Am JPathol 2009;175:448–57.

153. Lambert G, Chase MB, Dugi K, Bensadoun A,Brewer HB Jr, Santamarina-Fojo S. Hepaticlipase promotes the selective uptake of highdensity lipoprotein-cholesteryl esters via thescavenger receptor B1. J Lipid Res 1999;40:1294–303.

154. Blum CB, Deckelbaum RJ, Witte LD, Tall AR,Cornicelli J. Role of apolipoprotein E-containinglipoproteins in abetalipoproteinemia. J Clin In-vest 1982;70:1157–69.

155. Mahley RW, Huang Y, Weisgraber KH. Puttingcholesterol in its place: apoE and reverse cho-lesterol transport. J Clin Invest 2006;116:1226–9.

156. Linton MF, Atkinson JB, Fazio S. Prevention ofatherosclerosis in apolipoprotein E-deficientmice by bone marrow transplantation. Science1995;267:1034–7.

157. Hasty AH, Linton MF, Brandt SJ, Babaev VR,Gleaves LA, Fazio S. Retroviral gene therapy inApoE-deficient mice: ApoE expression in theartery wall reduces early foam cell lesion for-mation. Circulation 1999;99:2571–6.

158. Schaefer EJ, Gregg RE, Ghiselli G, Forte TM,Ordovas JM, Zech LA, Brewer HB, Jr. Familialapolipoprotein E deficiency. J Clin Invest 1986;78:1206–19.

159. Ghiselli G, Schaefer EJ, Gascon P, Breser HB Jr.Type III hyperlipoproteinemia associated withapolipoprotein E deficiency. Science 1981;214:1239–41.

160. Lohse P, Brewer HB 3rd, Meng MS, Skarlatos SI,LaRosa JC, Brewer HB Jr. Familial apolipoproteinE deficiency and type III hyperlipoproteinemiadue to a premature stop codon in the apoli-poprotein E gene. J Lipid Res 1992;33:1583–90.

HDL Nomenclature Special Report

Clinical Chemistry 57:3 (2011) 409

161. Vaisar T, Pennathur S, Green PS, Gharib SA,Hoofnagle AN, Cheung MC, et al. Shotgun pro-teomics implicates protease inhibition and com-plement activation in the antiinflammatoryproperties of HDL. J Clin Invest 2007;117:746–56.

162. Wolfrum C, Poy MN, Stoffel M. ApolipoproteinM is required for prebeta-HDL formation andcholesterol efflux to HDL and protects againstatherosclerosis. Nat Med 2005;11:418–22.

163. Christoffersen C, Jauhiainen M, Moser M, PorseB, Ehnholm C, Boesl M, et al. Effect of apoli-poprotein M on high density lipoprotein metab-olism and atherosclerosis in low density lipopro-tein receptor knock-out mice. J Biol Chem 2008;283:1839–47.

164. Axler O, Ahnstrom J, Dahlback B. An ELISA forapolipoprotein M reveals a strong correlation tototal cholesterol in human plasma. J. Lipid Res2007;48:1772–80.

165. Ahnstrom J, Axler O, Jauhiainen M, Salomaa V,Havulinna AS, Ehnholm C, et al. Levels of apo-lipoprotein M are not associated with the risk ofcoronary heart disease in two independent case-control studies. J Lipid Res 2008;49:1912–7.

166. Davidson WS, Silva RA, Chantepie S, Lagor WR,Chapman MJ, Kontush A. Proteomic analysis ofdefined HDL subpopulations reveals particle-specific protein clusters: relevance to antioxida-tive function. Arterioscler Thromb Vasc Biol2009;29:870–6.

167. Khatun I, Iqbal J, Wang M, Zeissig S, BlumbergR, Curiel D, et al. Phospholipid transfer activityof microsomal triglyceride transfer protein pro-motes assembly of phospholipid-rich apo-containing lipoproteins in mice. ArteriosclerThromb Vasc Biol. 2010;48 (abstr).

168. Collins LA, Mirza SP, Kissebah AH, Olivier M.Integrated approach for the comprehensive

characterization of lipoproteins from humanplasma using FPLC and nano-HPLC tandemmass spectrometry. Physiol Genomics 2010;40:208–15.

169. Hoofnagle AN, Heinecke J. Lipoproteomics:using mass spectrometry-based proteomics toexplore the assembly, structure, and functionof lipoproteins. J Lipid Res 2009;50:1967–75.

170. Green PS, Vaisar T, Pennathur S, Kulstad JJ,Moore AB, Marcovina S, et al. Combined statinand niacin therapy remodels the high-densitylipoprotein proteome. Circulation. 2008;118:1259–67.

171. Kontush A, Therond P, Zerrad A, Couturier M,Negre-Salvayre A, de Souza JA, et al. Prefer-ential sphingosine-1-phosphate enrichmentand sphingomyelin depletion are key featuresof small dense HDL3 particles: relevance toantiapoptotic and antioxidative activities.Arterioscler Thromb Vasc Biol 2007;27:1843–9.

172. Nilsson A, Duan RD. Absorption and lipoproteintransport of sphingomyelin. J Lipid Res 2006;47:154–71.

173. Duong PT, Collins HL, Nickel M, Lund-Katz S,Rothblat GH, Phillips MC. Characterization ofnascent HDL particles and microparticles formedby ABCA1-mediated efflux of cellular lipids toapoA-I. J Lipid Res 2006;47:832–43.

174. Shiflett AM, Bishop JR, Pahwa A, Hajduk SL.Human high density lipoproteins are platformsfor the assembly of multi-component innate im-mune complexes. J Biol Chem 2005;280:32578–85.

175. Nakamura Y, Kotite L, Gan Y, Spencer TA, Field-ing CJ, Fielding PE. Molecular mechanism ofreverse cholesterol transport: reaction of pre-beta-migrating high-density lipoprotein withplasma lecithin/cholesterol acyltransferase. Bio-

chemistry 2004;43:14811–20.176. Bolin DJ, Jonas A. Sphingomyelin inhibits the

lecithin-cholesterol acyltransferase reactionwith reconstituted high density lipoproteins bydecreasing enzyme binding. J Biol Chem 1996;271:19152–8.

177. de Souza JA, Vindis C, Hansel B, Negre-SalvayreA, Therond P, Serrano CV Jr, et al. Metabolicsyndrome features small, apolipoprotein A-I-poor, triglyceride-rich HDL3 particles with defec-tive anti-apoptotic activity. Atherosclerosis2008;197:84–94.

178. de Souza JA, Vindis C, Negre-Salvayre A, RyeKA, Couturier M, Therond P, et al. Small, denseHDL3 particles attenuates apoptosis in endothe-lial cells: Pivotal role of apolipoprotein A-I. J CellMol Med 2010;14:608–20.

179. Zerrad-Saadi A, Therond P, Chantepie S, Coutu-rier M, Rye K-A, et al. HDL3-Mediated inactiva-tion of LDL-associated phospholipid hydroperox-ides is determined by the redox status ofapolipoprotein A-I and HDL particle surface lipidrigidity: relevance to inflammation and athero-genesis. Arterioscler Thromb Vasc Biol 2009;29:2169–75.

180. Arsenault BJ, Lemieux I, Despres J-P, WarehamNJ, Stroes ESG, Kastelein JJP, et al. Comparisonbetween gradient gel electrophoresis and nu-clear magnetic resonance spectroscopy in esti-mating coronary heart disease risk associatedwith LDL and HDL particle size. Clin Chem 2010;56:789–96.

181. Asztalos BF, Roheim PS, Milani RL, Lefevre M,McNamara JR, Horvath KV, Schaefer EJ. Distri-bution of ApoA-I–containing HDL subpopula-tions in patients with coronary heart disease.Arterioscler Thromb Vasc Biol 2000;20:2670–6.

Special Report

410 Clinical Chemistry 57:3 (2011)