Development of a Homogeneous Assay for

Clinical Chemistry 57:1 57–65 (2011)

Lipids, Lipoproteins, and Cardiovascular Risk Factors

Development of a Homogeneous Assay for Measurement of Small Dense LDL Cholesterol Yasuki Ito,1* Miki Fujimura,1 Motoko Ohta,1 and Tsutomu Hirano2*

BACKGROUND: Plasma concentrations of small dense (sd)-LDL are associated with the prevalence of cardiovascular events. However, the special equipment and long assay times required for sd-LDL measurement have hindered its clinical application. Herein, we report development of a simple homogeneous assay for sd-LDL-cholesterol (C) adaptable to autoanalyzers. MATERIALS AND METHODS:

We identified suitable surfactants and phospholipases by screening for those selective for the sd-LDL fraction (d 1.044 –1.063 kg/L) and for the dissociation of other lipoproteins, including large buoyant LDL (lb-LDL). Principal characteristics of this assay were compared with ultracentrifugal isolation of LDL subfractions and with our previous heparin-magnesium precipitation assay for sd-LDL. We measured sd-LDL-C concentrations in 460 healthy, normolipidemic individuals.

RESULTS:

We used a polyoxyethylene benzylphenyl ether derivative to dissociate triglyceride-rich lipoproteins and HDLs, whereas sphingomyelinase proved most effective for dissociation of lb-LDL from LDL owing to the higher sphingomyelin content in the lb-LDL subfractions. A polyoxyethylene styrenephenyl ether derivative protected sd-LDL against the dissociative actions of sphingomyelinase and cholesterol oxidase/esterase during an initial incubation step. Next, polyoxyethylene alkyl ether dissociated sd-LDL-C and the cholesterol released from sd-LDL were subsequently measured by using cholesterol oxidase/esterase. The homogeneous method correlated excellently with ultracentrifugation for sd-LDL-C (y ⫽ 0.99x – 0.09, R2 ⫽ 0.91, n ⫽ 60) and exhibited within-run precision CVs ⬍1.1%. The distribution of sd-LDL-C was skewed, and the central 95% of sd-LDL-C concentrations ranged from 0.24 to 0.88 mmol/L (9.4 –34.0 mg/dL).

1

Reagent R&D Department, Denka Seiken Co., Ltd., Tokyo, Japan; 2 Department of Medicine, Division of Diabetes, Metabolism, and Endocrinology, Showa University School of Medicine, Tokyo, Japan. * Address correspondence to: Y.I. at Reagent R&D Department, Denka Seiken Co., Ltd., 3-4-2, Nihonbashi, Kayabacho, Chuo-ku, Tokyo, 103-0025, Japan. E-mail [email protected]. T.H. at Department of Medicine, Division of Diabetes, Metabolism, and Endocrinology, Showa University School of Medicine, Tokyo, Japan. E-mail [email protected].

CONCLUSIONS: The homogeneous assay allows reproducible measurement of sd-LDL-C within 10 min and appears promising in further investigations of the clinical significance of sd-LDL-C.

© 2010 American Association for Clinical Chemistry

LDL, an established atherogenic lipoprotein, can be fractionated into large buoyant (lb)3 and small dense (sd) particles based on size and density (1 ). An abundance of clinical evidence has shown that sd-LDL particles are more atherogenic than lb-LDL particles (2–5 ) and that a high sd-LDL cholesterol (C) concentration is closely associated with a high prevalence of cardiovascular disease (6 ). The selective measurement of the sdLDL-C concentration is thus important for evaluating the actual atherogenic risk of individuals. The sd-LDL is traditionally measured by ultracentrifugation (7 ) or gradient gel electrophoresis (GGE) (8 ). Yet these methods are both unsuitable for routine analysis, as each requires expensive equipment, complicated techniques, and long assay times. A recently developed method of nuclear magnetic resonance (NMR) imaging is capable of simultaneously determining the sizes and numbers of LDL particles (9 ), whereas HPLC for lipoprotein analysis enables the determination of lipid concentrations in various lipoprotein subfractions (10 ). However, both methods have limitations. HPLC is too laborious, time-consuming, and expensive for routine clinical use. The instrumentation and devices required for NMR are too costly for general clinical laboratories to feasibly acquire. We previously developed a simple precipitation method for sd-LDL-C quantification consisting of 2 steps: removal of apolipoprotein B (apoB)-containing sdLDL–free lipoproteins by precipitation with heparin and magnesium (Mg), followed by LDL-C measurement by the homogeneous method (11 ). Later, the

Received April 27, 2010; accepted October 26, 2010. Previously published online at DOI: 10.1373/clinchem.2010.149559 3 Nonstandard abbreviations: lb, large buoyant; sd, small dense; C, cholesterol; GGE, gradient gel electrophoresis; NMR, nuclear magnetic resonance; apoB, apolipoprotein B; PL, phospholipid; PLase, phospholipase; SMase, sphingomyelinase; SM, sphingomyelin; TOOS, N-ethyl-N-(2-hydroxy-3-sulfopropyl)-3methylaniline; 4-AA, 4-aminoantipyrine; PC, phosphatidylcholine; TG, triglyceride; TRL, TG-rich lipoprotein.

57

procedure was further simplified by using a filter to trap the aggregated lipoproteins with heparin-Mg (12 ). Yet even with this simplification, the technique still required an offline sample pretreatment that hindered the smooth integration of this method into general clinical examination. Therefore, there has been need for the development of a fully automated assay for sd-LDL-C. Specific interactions of surfactants with lipoproteins have been successfully used to develop direct lipoprotein-C assays in which non-LDL lipoproteins are dissociated with a specific surfactant to allow the cholesterol to be accessed as a nonchromogenic product. A homogeneous assay for LDL-C was established by eliminating non-LDL lipoproteins with a specific surfactant (13 ). The surface of lipoproteins consists of various phospholipids (PLs), and phospholipase (PLase) can change the composition and property of lipoproteins (14 ). Sphingomyelinase (SMase) reportedly aggregates and fuses LDL particles (15–17 ). Noting that sphingomyelin (SM) is more abundant in larger and less dense LDL particles (14 ), we speculated that SMase might preferentially dissociate lb-LDL, thus enabling the detection of the remaining LDL (which is sd-LDL) by a general cholesterol assay. In this study, we screened 2 different surfactants: one that dissociated non-LDL lipoproteins and another that protected sdLDL particles against the actions of SMase and cholesterol esterase/oxidase. Finally, by combining these 2 surfactants with SMase, we established a novel homogeneous assay for sd-LDL-C. Materials and Methods MATERIALS

Cholesterol esterase (EC3.1.1.13 Pseudomonas sp.) was obtained from Asahi Kasei Pharma, cholesterol oxidase (EC1.1.3.6 microorganism) and peroxidase (EC1.11.1.7 Horseradish) from Toyobo, and catalase (EC1.11.1.6 Corynebacterium glutamicum) from Roche Diagnostics. N-Ethyl-N-(2-hydroxy-3-sulfopropyl)-3methylaniline (TOOS) and Good’s buffer (PIPES) were supplied from Dojindo Laboratories, 4-aminoantipyrine (4-AA) from Nittobo, and sodium azide from Nacalai Tesque. Surfactants (a polyoxyethylene benzylphenyl ether derivative, a polyoxyethylene styrenephenyl ether derivative, and polyoxyethylene alkyl ether) were obtained from Kao Corporation. PLase A2, C, D, lysophospholipase, glycerophospholipase, and SMase derived from Streptomyces sp. were obtained from Asahi Kasei Pharma. MEASUREMENTS

We measured total cholesterol, LDL cholesterol, and PL by use of commercially available test kits [T-CHO, 58 Clinical Chemistry 57:1 (2011)

LDL-EX(N), and PL-S, respectively; Denka Seiken]. We determined the LDL size phenotype by 2%–16% polyacrylamide GGE according to the method of Krauss and colleagues (2 ). We measured sd-LDL-C by use of a test kit based on the precipitation method with filtration (12 ) (sd-LDL-C “Seiken”; Denka Seiken) and SM according to the method of Hojjati and Jiang (18 ). Phosphatidylcholine (PC) was determined by subtracting SM from PL. SEPARATION OF LDL SUBFRACTION BY GEL FILTRATION

We separated serum samples into lipoprotein fractions by gel filtration chromatography using Superose 6HR 10/30 columns (Pharmacia). Each sample (0.2 mL) was loaded, and each lipoprotein fraction was eluted with PBS buffer at a flow rate of 0.5 mL/min. The column effluent was collected in 0.5-mL fractions, and the cholesterol concentration of each was measured by enzymatic assay (13 ). SEPARATION OF LIPOPROTEINS AND LDL SUBFRACTIONS BY ULTRACENTRIFUGATION

We separated triglyceride (TG)-rich lipoproteins (TRLs) (d ⬍1.019 kg/L) and HDL (d 1.063–1.21 kg/L) from serum by sequential flotation in an ultracentrifuge (Himac CS120GX; Hitachi Koki) and S100AT6 rotor (Hitachi Koki) according to the method of Havel et al. (19 ). We separated lb-LDL (d 1.019 –1.044 kg/L) and sd-LDL (d 1.044 –1.063 kg/L) according to the method of Havel et al., with modifications. The method of ultracentrifugation for LDL subclasses was as follows: 2 mL serum was mixed with 1 mL of d 1.045 kg/L NaBr solution and centrifuged at 387 000g (max) for 5 h. After removal of TRL, d 1.094 kg/L NaBr solution was mixed and centrifuged at 604 000g for 23 h. This centrifugation resulted in spontaneous gradient generation inside the tube, and 20 LDL subfractions (0.1 mL each) were collected by the fraction collector. The density of each fraction was determined by DA500 (Kyoto Electronics Manufacturing), and we confirmed that the isolated LDL fraction corresponded to 1.019 – 1.063 kg/L. SCREENING OF SURFACTANTS AND PLase

We used the following 2 reagents to select the specificity of surfactants and PLase and evaluate assay performance. Reagent 1 consisted of 600 U/L cholesterol esterase, 500 U/L cholesterol oxidase, 1200 KU/L catalase, 2 mmol/L N-ethyl-N-(2-hydroxy3-sulfopropyl)-3-methylaniline, sodium salt, and dihydrate in PIPES buffer (50 mmol/L, pH 7.0). Reagent 2 consisted of 5 KU/L peroxidase, 4 mmol/L 4-aminoantipyrine, 0.05% (wt/vol) sodium azide, and 1% (wt/vol) polyoxyethylene alkyl ether, in piperazineN,N⬘-bis(2-ethanesulfonic acid) buffer. A 1500-␮L al-

Homogeneous Assay for Small Dense LDL

iquot of reagent 1 was added to 3 ␮L sample, and the solution was incubated at 37 °C for 5 min (step 1). Next, 50 ␮L of reagent 2 was added, and the reaction mixture was incubated at 37 °C for 5 min (step 2). In screening, some lipoproteins were dissociated to water and oxygen by actions of cholesterol esterase/ oxidase and catalase to form a nonchromogenic product in step 1. The lipoprotein left over from step 1 conferred a purple-red color with the coupler in the presence of peroxidase in step 2, and cholesterol was measured spectrophotometrically at dual wavelengths of 600 nm (main) and 700 nm (subsidiary). This assay was performed in an autoanalyzer (Hitachi 7180). Screening of surfactant A. In screening for surfactant A, various surfactants were added to reagent 1 followed by measurement of each lipoprotein isolated by ultracentrifugation. These lipoprotein measurements (made using total cholesterol reagents) were compared for the various surfactants investigated. Screening of SMase. To determine an optimal PLase, several PLases were added to reagent 1 including surfactant A. Measurements of sd-LDL-C were performed on several sera and compared with sd-LDL-C as measured by the precipitation method. Screening of surfactant B. Surfactant B was chosen by the following approach: various surfactants were added to reagent 1 along with surfactant A and SMase. Subsequent measurements were made of lb-LDL-C and sdLDL-C purified by ultracentrifugation using total cholesterol reagents, and these measurements were compared for the various surfactants investigated. LDL SIZE MEASUREMENT BY ELECTRON MICROSCOPY

Serum lipoprotein particle sizes before and after treatment with the reagents of the sd-LDL-C homogeneous method were observed by electron microscopy according to the method of Nakata et al. (20 ), with modifications. Grids coated with collodion and carbon were made hydrophilic by 0.1% (wt/vol) poly-L-lysine and floated on a drop of sample suspension for 5 min. After removing excess fluid by dabbing the edge of each grid with a piece of paper, the grids were floated on drops of 2% phosphotungstic acid for 2 min and immediately drained to near dryness on absorbed filter paper. Lipoproteins were stained with osmium. REFERENCE INTERVAL OF sd-LDL-C DETERMINED BY THE HOMOGENEOUS ASSAY

We determined the reference interval of sd-LDL-C by measuring serum samples from 460 normolipidemic healthy individuals [LDL ⬍3.64 mmol/L (140 mg/dL), TG ⬍1.65 mmol/L (150 mg/dL), and HDL-C ⬎1.04 mmol/L (40 mg/dL) according to the Japan Atheroscle-

rosis Society guidelines] (21 ). Written informed consent was obtained from all individuals, and the local ethics committee approved the experimental protocol. Results In total, we screened 174 surfactants for their ability to dissociate TRL and HDL. Through this screening, polyoxyethylene benzylphenyl ether derivative (surfactant A) was found to be the most effective agent for selectively dissociating TRL and HDL without affecting LDL. Surfactant A specifically dissociated TRL and HDL, while LDL remained intact, irrespective of the presence of hypertriglyceridemia (see Supplemental Fig. 1, which accompanies the online version of this article at http://www.clinchem.org/content/vol57/ issue1). We examined compositional changes in LDL with increased density. Fig. 1 shows the cholesterol concentration (A) and PC and SM concentrations (B) in LDL subspecies fractionated by the ultracentrifugation. The cholesterol, PC, and SM concentrations all increased, reached peak values at fraction 11, and then decreased as the LDL density increased (11–19 ). The SM/LDL particle ratio, estimated by the apoB concentration, consistently fell as the LDL density rose. The SM/PC ratio plateaued between fractions 1 and 11, then decreased as the LDL density increased (11–19 ). The densities corresponding to the LDL subfractions are shown in Fig. 1D. These results suggested SM as a target lipid for separating LDL subspecies. Fig. 2 demonstrates the linear regression curves between sd-LDL-C concentrations determined by the precipitation method and cholesterol concentration after treatment with various PLases in presence of surfactant A (the homogeneous method). Serum samples were obtained from healthy volunteers. Cholesterol concentrations treated with PLase A2, PLase C, PLase D, lysophospholipase, and glycerophospholipase were weakly correlated with sd-LDL-C determined by the precipitation method (R2 ⫽ 0.13– 0.67). On the other hand, the cholesterol concentration treated with SMase was highly correlated with sd-LDL-C determined by the precipitation method (R2 ⫽ 0.895). The SMase concentration was set at 600 U/L since a decline and a plateau of the cholesterol reaction during the rise of the SMase concentration to 500 U/L were observed (see online Supplemental Fig. 2). Ultimately, SMase was selected as the PLase most effective in dissociating lbLDL and preserving sd-LDL. It was also found, however, that homogeneous assay results after treatment with surfactant A and SMase still had a nonzero intercept against the sd-LDL-C concentration as determined by the precipitation method [i.e., the intercept was 0.3 mmol/L (12.3 mg/dL)]. This nonzero intercept Clinical Chemistry 57:1 (2011) 59

A

6 4 2

CHO Density

0

1.07 1.06 1.05 1.04 1.03 1.02 1.01 1 00 1.00

B 1.5 PC SM

1.0

0.5

00 0.0

1 3 5 7 9 11 13 15 17 19

C 2500

1 3 5 7 9 11 13 15 17 19 0.40

D

Fraction

Density,

Fraction

Density,

11 12 13 14 15 16 17 18 19 20

1.042 1.043 1.045 1.047 1.049 1.050 1.052 1.055 1.057 1.059

g/L

2000

0.30

1500 0.20

1000 500

SM/LDL SM/PC

0

0.10 0.00

1 3 5 7 9 11 13 15 17 19 Ultracentrifugation fraction no.

1 2 3 4 5 6 7 8 9 10

1.021 1.032 1.033 1.034 1.035 1.036 1.037 1.038 1.039 1.041

g/L

Fig. 1. Cholesterol (CHO) concentration (A) and SM and phosphatidylcholine (PC) concentrations (B) in isolated LDL subfractions by gradient ultracentrifugation (density 1.019 –1.063 kg/L). (C), The SM/LDL particle and the SM/PC ratio in isolated LDL subfractions. The LDL particle number is estimated by the apoB concentration. (D), Corresponding densities for each LDL subfraction.

suggested that the specificity of the homogeneous method with surfactant A and SMase was not sufficient to accurately determine sd-LDL-C values. Fig. 3A shows the reactivity of narrow-cut lb-LDL (fraction 3, d 1.033 kg/L) and narrow-cut sd-LDL (fraction 15, d 1.059 kg/L) with the total-cholesterol assay after treatment by surfactant A (0.07% wt/vol) with and without SMase (600 U/L), where the absorbance without surfactant A was set as 100%. Both lbLDL-C and sd-LDL-C reacted well with the totalcholesterol assay in the absence of SMase. In contrast, only 3% of the lb-LDL-C and 18% of the sd-LDL-C reacted with the cholesterol assay in the presence of SMase. After screening for a second surfactant (surfactant B) to protect sd-LDL against the reaction of SMase, we found that a polyoxyethylene styrenephenyl ether derivative (Kao Corp.) can protect sd-LDL against the actions of SMase and cholesterol esterase/oxidase. Fig. 3B shows the reactivity of lb-LDL (fraction 3) and sdLDL (fraction 15) with the total-cholesterol assay without and with surfactant B (0.12% wt/vol), in the presence of surfactant A and SMase. In the absence of surfactant B, both lb-LDL and sd-LDL were dissociated by the action of SMase, and only 3% of the lb-LDL-C 60 Clinical Chemistry 57:1 (2011)

and 18% of the sd-LDL-C were detected by the totalcholesterol assay when the absorbance without surfactant A and SMase was set at 100%. In the presence of surfactant B, sd-LDL was strongly protected from the actions of surfactant A and SMase, and ⬎80% of sdLDL-C could be detected. With the surfactant B treatment, however, the reactivity of lb-LDL-C was only partially increased and most of the lb-LDL (90%) remained undetected by the cholesterol assay. The principle of this homogeneous assay for sdLDL-C is shown in Fig. 4. Surfactant A reacts with TRL and HDL, and the cholesterol in these lipoproteins is dissociated by the action of cholesterol oxidase/esterase and catalase. SMase mainly reacts with lb-LDL, whereas surfactant B protected sd-LDL from the actions of SMase and cholesterol-oxidase/esterase reaction (R1). Thus, the sd-LDL-C that escapes from the action of these enzymes can be measured by the standard total-cholesterol assay (R2). Figs. 5A and B shows the reaction of the homogeneous method with ultracentrifugally isolated LDL obtained from serum of the lb-LDL (A) and sd-LDL (B) phenotype determined by GGE (19 ). The sd-LDL homogeneous method selectively elicited a reaction with the sd-LDL (fractions 13–20), whereas the LDL homo-


Fig. 2. Linear regression curves between the sd-LDL-C values determined by the precipitation method [Hirano et al. (12 )] and sd-LDL-C values determined by the homogeneous method with surfactant A and various phospholipases. Among the homogeneous methods using phospholipases, that using sphingomyelinase exhibited the highest correlation with sd-LDL-C, as determined by the precipitation method.

A

B

100%

80% Reactiviity

Reactiviity

80%

100%

60% 40% 20%

60% 40% 20%

0%

0% Fraction 3 Fraction 15 (lb-LDL) (sd-LDL)

Fraction 3 Fraction 15 (lb-LDL) (sd-LDL)

SMase(−) SMase(+)

Surfactant B(−) Surfactant B(+)

Fig. 3. (A), The effect of SMase on the reactivity of narrow-cut lb-LDL (ultracentrifugation fraction 3, d 1.033 kg/L) and narrow-cut sd-LDL (fraction 15, d 1.059) with the total-cholesterol assay in the presence of surfactant A. The absorbance without surfactant A was set at 100%. (B), Effect of surfactant B on the reactivity (% absorbance) of lb-LDL (fraction 3, d 1.033 kg/L) and sd-LDL (fraction 15, d 1.049) with the total-cholesterol assay system in the presence of surfactant A and SMase. The absorbance without surfactant A and SMase was set at 100%.

geneous elicited a reaction with the total LDL fraction. The homogeneous assay sd-LDL-C results for 60 samples from healthy volunteers were compared with those obtained by ultracentrifugation. Serum samples were obtained from these individuals who had variable serum lipid levels [TG, 0.3–5.2 mmol/L (30 – 472 mg/ dL), LDL-C, 1.6 –5.2 mmol/L (63–199 mg/dL), and HDL-C, 0.9 –2.5 mmol/L (35–97 mg/dL)]. The sdLDL-C concentrations determined by the homogeneous method and ultracentrifugation were excellently correlated in all subjects (y ⫽ 0.99x – 0.09, R2 ⫽ 0.91 Pearson correlation). No significant differences were found between the mean (standard deviation) sdLDL-C concentrations determined by the homogeneous method and by ultracentrifugation in the overall study population [0.81 (0.06) vs 0.91 (0.06) mmol/L, or 31.2 (2.2) vs 35.0 (2.1) mg/dL, respectively; P ⫽ 0.220] or in the dyslipidemic subjects [TG, ⬎1.65 mmol/L (150 mg/dL); LDL-C, ⬎3.64 mmol/L (140 mg/dL); or HDL-C, ⬍1.04 mmol/L (40 mg/dL); 1.06 (0.08) vs 1.15 (0.08) mmol/L, or 40.6 (3.1) vs 43.9 (3.0) mg/dL, respectively; P ⫽ 0.445]. Fig. 6 shows the Deming fit for the sd-LDL-C values determined by the homogeneous method and ultracentrifugation. The Deming fit plot analysis revealed no significant proportional difference Clinical Chemistry 57:1 (2011) 61

TRL lb-LDL

Step1

sd-LDL

Surfactant A

Sphingomyelinase

Dissociate

Dissociate

Cholesterol

Surfactant B Protect

Purple-red color

Measurement

2H2O + O2

2H2O + O2

H2 O2

Dissociate

Cholesterol esterase Cholesterol oxidase Catalase

Cholesterol esterase Cholesterol oxidase Catalase

Step2

Surfactant A

Cholesterol

Cholesterol

2H2O + O2

HDL

Cholesterol

Cholesterol esterase Cholesterol oxidase Peroxidase

sd-LDL

Fig. 4. The simple principle of the homogeneous assay for sd-LDL-C. Surfactant A (polyoxyethylene benzylphenyl ether derivative) reacts with TRLs and HDL, and cholesterol in these lipoproteins is eliminated by the action of cholesterol oxidase/esterase and catalase in step 1. Sphingomyelinase specifically reacts with lb-LDL, while surfactant B (polyoxyethylene styrenephenyl ether derivative) protects sd-LDL from the actions of sphingomyelinase and cholesterol oxidase/esterase (R1). The sd-LDL-C that escapes via the action of these enzymes is measured by the standard cholesterol assay in step 2.

between the 2 methods (slope 1.04, 95% CI 0.94 –1.13, SE 0.48, P ⫽ 0.41), but a slight constant negative intercept was found for the homogeneous method in comparison with ultracentrifugation [intercept ⫺0.13 mmol/l (⫺5.0 mg/dL), 95% CI ⫺0.22 to ⫺0.05, SE 0.042, P ⬍ 0.005]. The homogeneous sd-LDL-C method also exhibited a close association with sd-LDL-C determined by the precipitation method (y ⫽ 0.99x – 0.04, R2 ⫽ 0.82) (see online Supplemental Fig. 3). Electron microscopic observations confirmed that only LDL particles with a diameter of 15–20 nm (average 17 nm) were left after the incubation of plasma with surfactants A and B and SMase (see online Supplemental Fig. 4). We evaluated the basic characteristics of the homogeneous assay for sd-LDL-C (see online Supplemental Table 1). To investigate for possible interferences, we used Interference Check A Plus (Sysmex) made from human blood. The precision for the measurement of sd-LDL-C was determined by duplicate determinations, in 2 runs/day for 21 days (total n ⫽ 84). The CVs for within-run precision were ⬍1.1%, and total CVs were ⬍2.2%. The assay did not show evidence of interference with 2.52 mmol/L (50 mg/dL) ascorbic acid, 5.0 g/L hemoglobin (hemolysate), 0.51 62 Clinical Chemistry 57:1 (2011)

mmol/L (30 mg/dL) conjugated bilirubin, or 0.51 mmol/L (30 mg/dL) unconjugated bilirubin. When we examined the effect of triglyceride on the assay by adding isolated TRL from severe hypertriglyceridemic subjects, we found that plasma TG concentrations ⬍11 mmol/L (⬍1000 mg/dL) did not affect the sd-LDL-C measurement (see online Supplemental Fig. 5). Mean sd-LDL-C using serum vs plasma for the sample showed a negligible difference [1.00 ⫹ 0.04 vs 0.93 ⫹ 0.04 mmol/L (38.6 ⫹ 1.8 vs 35.9 ⫹ 1.7 mg/dL), respectively, n ⫽ 53, P ⫽ 0.266], and the serum samples were stable until the fourth day of incubation at 4 °C. The distribution of sd-LDL-C levels was skewed in normolipidemic individuals. The central 95% of sdLDL-C concentrations ranged from 0.24 to 0.88 mmol/L (9.4 –34.0 mg/dL) (see online Supplemental Fig. 6). Discussion In previous experiments with a homogeneous method (LDL-EX; Denka Seiken) (13 ) using the surfactant to dissociate TRL and HDL, selectively measuring narrow-cut LDL in a density range of 1.019 –1.063 kg/L was successful. On that basis, we believed it would be possible to establish a homogeneous method for sd-

Phenotype A 1.6 1.2 0.8 0.4 0

1

3

5

Choleesterol (mmol/L)

B

7

9

11 13 15 17 19

Phenotype B

2.0 n = 60 y = 1.04x – 0.13

1.5 1.0 0.5 0.0

Identical Deming fit

0.0

0.5

1.0

1.5

2.0

Ultracentrifugation method (mmol/L)

2.5 2 1.5

4

1

3

0.5 0

1

3

5

7

9

11 13 15 17 19

Ultracentrifugation fraction no. sd-LDL-C

LDL-C

Fig. 5. Reaction of the homogeneous method with ultracentrifugally isolated LDL subfractions obtained from the sera of the lb-LDL (A) and sd-LDL (B) phenotype determined by GGE [Nichols et al. (8 )].

LDL by finding a substance effective in dissociating lbLDL. Liu et al. (14 ) reported that SM or phosphatidylinositol and phosphatidylserine were more abundant in lb-LDL than in sd-LDL. Kontush et al. (22 ) found that this abundance of sphingomyelin diminished in parallel with a progressive increase in HDL density. Similarly, we observed that the SM content of LDL particles and the SM/PC ratio decreased as the LDL particle density rose. Several articles have demonstrated that SMase aggregates and fuses LDL particles (15–17 ). We speculated that SMase might preferentially decompose lb-LDL particles containing abundant SM, while imposing only minimal effects on sd-LDL particles containing smaller amounts of SM. In tests to determine which PLases were most suitable for sd-LDL measurement, we found that the homogeneous assay system using SMase correlated most closely with the established method for sd-LDL-C precipitation. SMase dissociated lb-LDL almost completely and sd-LDL partially, a finding that could have been predicted by the levels of SM in lb-LDL and sd-LDL. Given that SMase would be expected to partially dissociate the sd-LDL particles, we knew that another reagent would be required to protect the sd-LDL particles against the action of SMase. To find the best candidate, we

Standardized residual

Cholesteroll (mmol/L)

A

Homo ogeneous metthod (mmol/L)


2 1 0 −1 −2 −3 −4 0.0

0.5

1.0

1.5

2.0

Estimated “true” sd-LDL-C (mmol/L) Fig. 6. Simple correlation between the homogeneous method and ultracentrifugation for the sd-LDL-C values (A). Standardized residuals of Deming regression fit for comparison of sd-LDL-C values between the homogeneous method and ultracentrifugation (B). ⫻, Dyslipidemic individuals (TG ⬎1.65 mmol/L, LDL-C ⬎3.64 mmol/L or HDL-C ⬍1.04 mmol/L); E, normolipidemic individuals.

screened the effectiveness of various surfactants in protecting sd-LDL against the actions of SMase, cholesterol oxidase/esterase, and catalase in step 1. On the basis of these investigations, we selected polyoxyethylene styrenephenyl ether derivative as surfactant B. The reactivity of sd-LDL-C with the cholesterol assay in the presence of surfactant B was strongly protected, rising from 18% to 82%. With this finding, it appeared that surfactant B was critical for the protection of sd-LDL particles against dissociation with SMase and cholesterol oxidase/esterase. The precise mechanisms for surfactant B–induced protection of sd-LDL particles in step 1 remain unknown. The new homogeneous method gave good agreement with the ultracentrifugation method, a standard procedure for the measurement of sd-LDL (d 1.044 – 1.063 kg/L). Specifically, a regression equation, y ⫽ Clinical Chemistry 57:1 (2011) 63

0.99x – 0.09, R2 ⫽ 0.91, was obtained when comparing the homogeneous method with ultracentrifugation, a better correlation than obtained with our previous precipitation method (y ⫽ 1.05x ⫹ 1.0, R2 ⫽ 0.78) (11 ). In addition, the absolute values of sd-LDL-C were almost identical between the ultracentrifugation and homogeneous methods. In these experiments, the homogeneous method was also confirmed to have specificity for the sd-LDL subfractions obtained from 2 different LDL phenotypes, patterns A and B. Electron microscopic observations also supported the finding that the homogeneous system selectively dissociated other lipoproteins while excluding sd-LDL particles. All of these observations strongly corroborate the validity of the new homogeneous assay for selectively determining sd-LDL-C and dissociating lb-LDL-C. However, precise mechanisms to explain why the homogeneous method selectively decomposes d ⬍1.044 kg/L lipoproteins have not been identified. Compositional changes of sd-LDL particles, such as a lower SM, PL, or cholesterol (23 ), might occur around densities of 1.044 kg/L. Further study will be required to elucidate this important issue. The distribution of sd-LDL-C concentrations determined by the homogeneous method was skewed. The mean sd-LDL-C in normolipidemic individuals was approximately 0.5 mmol/L, or almost identical to the value measured earlier by the precipitation method (24 ). As a technique now proven to enable the mea-

surement of sd-LDL-C in a fully automated manner with a short assay time, our homogeneous method is expected to be helpful in advancing understanding of the clinical significance of sd-LDL-C in various clinical studies.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article. Authors’ Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest: Employment or Leadership: None declared. Consultant or Advisory Role: None declared. Stock Ownership: None declared. Honoraria: None declared. Research Funding: This study was supported by Denka Seiken Co., Ltd., and Showa University. Expert Testimony: None declared. 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. Acknowledgments: The authors are indebted to Dr. Katsuyuki Nakajima at Gumma University for his critical comments and suggestions.

References 1. Berneis KK, Krauss RM. Metabolic origins and clinical significance of LDL heterogeneity. J Lipid Res 2002;43:1363–79. 2. Austin MA, Breslow JL, Hennekens CH, Buring JE, Willet WC, Krauss RM. Low-density lipoprotein subclass patterns and risk of myocardial infarction. JAMA 1988;260:1917–21. 3. Austin MA, King MC, Viranizan KM, Krauss RM. Atherogenic lipoprotein phenotype: a proposed genetic marker for coronary heart disease risk. Circulation 1990;82:495–506. 4. Lamarche B, Tchernof A, Moorjani S, Cantin B, Dagenais GR, Lupien PJ, et al. Small, dense lowdensity lipoprotein particles as a predictor of the risk of ischemic heart disease in men. Circulation 1997;95:69 –75. 5. Gardner CD, Fortmann SP, Krauss RM. Association of small low-density lipoproteins particles with the incidence of coronary artery disease in men and women. JAMA 1996;276:875– 81. 6. St-Pierre AC, Cantin B, Dagenais GR, Mauriege P, Bernard P-M, Despres JP, et al. Low-density lipoprotein subfractions and the long-term risk of ischemic heart disease in men. Arterioscler Thromb Vasc Biol 2005;25:553–9. 7. Swinkels DW, Hak-Lemmers HLM, Demacker PNM. Single spin density gradient ultracentrifugation method for the detection and isolation of light and heavy low density lipoprotein subfractions. J Lipid Res 1987;28:1233–9.

64 Clinical Chemistry 57:1 (2011)

8. Nichols AV, Krauss RM, Musliner TA. Nondenaturing polyacrylamide gradient gel electrophoresis. Methods Enzymol 1986;128:417–33. 9. Otvos JD, Jeyarajah EJ, Bennett DW, Krauss RM. Development of a proton nuclear magnetic resonance spectroscopic method for determining plasma lipoprotein concentrations and subspecies distributions from a single, rapid measurement. Clin Chem 1992;38:1632– 8. 10. Okazaki M, Usui S, Hosaki S. Analysis of plasma lipoproteins by gel permeation chromatography. In: Rifai N, Warnick GR, Dominiczak MH, eds. Handbook of lipoprotein testing. 2nd ed. Washington (DC): AACC Press; 2000. p 647– 69. 11. Hirano T, Ito Y, Saegusa H, Yoshino G. A novel and simple test method for quantification of small, dense LDL. J Lipid Res 2003;44:2193–201. 12. Hirano T, Ito Y, Yoshino G. Measurement of small dense low-density lipoprotein particles. J Atheroscler Thromb 2005;12:67–72. 13. Okada M, Matsui H, Ito Y, Fujiwara A, Inano K. Low-density lipoprotein cholesterol can be chemically measured: a new superior method. J Lab Clin Med 1998;132:195–201. 14. Liu M, Krul ES, Subbaiah PV. Effect of apoprotein B conformation on the activation of lysolecithin acyltransferase and lecithin: cholesterol acyltransferase. Studies with subfractions of low density lipoproteins. J Biol Chem 1992; 267:5139 – 47.

15. Oörni K, Hakala JK, Annila A, Ala-Korpela M, Kovanen PT. Sphingomyelinase induces aggregation and fusion, but phospholipase A2 only aggregation, of low density lipoprotein (LDL) particles: two distinct mechanisms leading to increased binding strength of LDL to human aortic proteoglycans. J Biol Chem 1998;273: 29127–34. 16. Xu X-X, Tabas I. Sphingomyelinase enhances low density lipoprotein uptake and ability to induce cholesteryl ester accumulation in macrophages. J Biol Chem 1991;266:24849 –58. 17. Pentika¨inen MO, Lehtonen EMP, Kovanen PT. Aggregation and fusion of modified low density lipoprotein. J Lipid Res 1996;37:2638 – 49. 18. Hojjati MR, Jiang XC. Rapid, specific, and sensitive measurements of plasma sphingomyelin and phosphatidylcholine. J Lipid Res 2006;47: 673– 6. 19. Havel RJ, Eder HA, Bragdon JH. The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest 1955;34:1345–53. 20. Nakata S, Petrie BL, Calomeni EP, Estes MK. Electron microscopy procedure influences detection of rotaviruses. J Clin Microbiol 1987;25: 1902– 6. 21. Teramoto T, Sasaki J, Ueshima H, Egusa G, Kinoshita M, Shimamoto K, et al. Risk factors of atherosclerotic diseases: executive summary of


Japan Atherosclerosis Society (JAS) guideline for diagnosis and prevention of atherosclerosis cardiovascular diseases for Japanese. J Atheroscler Thromb 2007;14:267–77. 22. Kontush A, Therond P, Zerrad A, Couturier M, Ne´gre-Salvayre A, de Souza JA, et al. Preferential sphingosine-1-phosphate enrichment and sphin-

gomyelin depletion are key features of small dense HDL3 particles: relevance to antiapoptotic and antioxidative activities. Arterioscler Thromb Vasc Biol 2007;27:1843–9. 23. McNamara JR, Small DM, Li Z, Schaefer EJ. Differences in LDL subspecies involve alterations in lipid composition and conformational

changes in apolipoprotein B. J Lipid Res 1996; 37:1924 –35. 24. Hirano T, Nohtomi K, Sato Y, Kamata K, Ito Y. Small dense LDL-cholesterol determined by a simple precipitation assay for screening familial combined hyperlipidemia. Atherosclerosis 2009;205: 603–7.

Clinical Chemistry 57:1 (2011) 65