Clinica Chimica Acta 436 (2014) 348–350

Contents lists available at ScienceDirect

Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim

Immunoprecipitation of apolipoprotein B-containing lipoproteins for isolation of HDL particles John H. Contois ⁎, Andre L. Albert, Rae-Anne Nguyen Sun Diagnostics, LLC, 60 Pineland Drive, Brunswick Hall, Suite 322, New Gloucester, ME 04260, USA

a r t i c l e

i n f o

Article history: Received 17 April 2014 Received in revised form 17 June 2014 Accepted 18 June 2014 Available online 30 June 2014 Keywords: High density lipoprotein Immunoprecipitation Lipoproteins Cholesterol

a b s t r a c t Background: Immunoprecipitation (IP) of non-HDL particles with antisera provides the simplest and most specific method available for the separation of HDL. We compared the LipoSep™ IP reagent with the dextran sulfate/MgCl2 precipitation method (DS). Methods: The IP reagent (200 μl) was added to an equal volume of serum, vortexed, incubated for 10 min at room temperature, and microcentrifuged at 12,000 rpm for 10 min. Results: Equal volumes of a sample and the IP reagent precipitated apoB to 3.0 g/l without the coprecipitation of HDL. HDL-C measured in the supernatant after IP (Y) gave excellent agreement to DS precipitation (X) with a slope of 1.01, an intercept of 0.070 mmol/l (2.7 mg/dl), and a correlation of 0.99 (n = 118; apoB 0.16–2.11 g/l). However, DS failed in most samples with moderate to elevated triglycerides. At triglyceride concentrations from 2.86 to 23.63 mmol/l (253–2091 mg/dl) the initial success rate was 65.4% for IP, while DS successfully precipitated only 5.8%. Success rate on repeat with additional reagent and/or sample dilution gave a success rate of 86.5% for IP and 40.4% for DS. Conclusion: The IP reagent and protocol is a simple, effective and highly specific tool for isolating HDL particles in human serum and is effective with high triglyceride samples. © 2014 Elsevier B.V. All rights reserved.

1. Introduction The characterization of high density lipoproteins (HDL) is currently the focus of intense research. HDL particles play an important role in reverse cholesterol transport (RCT) and protection from coronary heart disease (CHD). Although RCT is presumed to be the primary mechanism for HDL's protective effect, HDL also have antiinflammatory, antioxidant, and antithrombotic activities, and appear to promote healthy endothelial function [1]. The concept of “dysfunctional HDL”, HDL particles that are somehow altered and no longer protective against CHD, is relatively new. Clinicians have long noticed the paradox that some individuals with very high HDL cholesterol concentrations develop CHD in the absence of obvious risk factors. Recent data have implicated certain proteins associated with these dysfunctional HDL [2]. In addition to HDL function, new classes of pharmaceuticals, including CETP inhibitors, are being developed to raise HDL concentrations. The challenge for researchers is to quantitatively isolate HDL particles for the measurement of cholesterol and/or other components that may explain function or better monitor therapeutic efficacy.

Traditional approaches such as electrophoresis and ultracentrifugation are labor intensive, technically demanding, and may alter particle composition and structure. Chemical methods such as dextran sulfate/ magnesium chloride (DS/MgCl2) precipitation may not completely remove apolipoprotein B (apoB)-containing lipoproteins or adequately capture all HDL subclasses [3]. Homogeneous methods for HDL cholesterol measurement have proven to be inadequate in patients with certain lipoprotein abnormalities [4]. Immunoprecipitation shares the positive attributes of chemical precipitation methods such as ease of use, but because of the specificity afforded by antibodies, also allows quantitative separation of HDL particles from apoB lipoproteins, without altering lipoprotein particle composition. Now that immunoprecipitation reagent for the separation of HDL is commercially available, we see uses as a reference method for HDL cholesterol measurement, a secondary method for HDL cholesterol measurement in samples with elevated triglycerides or where direct HDL cholesterol measurement is suspect, and a research tool to quantitatively separate HDL particles for functional assays and the measurement of HDL-specific analytes. 2. Methods

⁎ Corresponding author. Tel.: +1 207 926 1125; fax: +1 207 926 1126. E-mail address: [email protected] (J.H. Contois).

http://dx.doi.org/10.1016/j.cca.2014.06.017 0009-8981/© 2014 Elsevier B.V. All rights reserved.

The IP reagent was prepared from delipidated and stabilized goat anti-apoB sera (LipoSep IP™, Sun Diagnostics). Cholesterol (Wako Chemicals), apolipoprotein AI (apoAI) and apoB (laboratory developed

J.H. Contois et al. / Clinica Chimica Acta 436 (2014) 348–350

immunoturbidimetric assays) were measured using a Cobas Fara II analyzer. Because of the importance of eliminating apoB-containing lipoprotein particles the limit of blank (LOB) and limit of detection (LOD) were determined as recommended by CLSI for the apoB assay [5]. To assess the limit of the blank (LOB) apoB was measured in 50 replicates of 0.9% saline over 5 days. LOB was calculated as mean + (c ∗ SD); where c = 1.645 / (1 − (1 / (4 ∗ df))) [5]. The limit of detection (LOD) was calculated from 49 replicate apoB measurements over 5 days of a diluted serum pool with a concentration ~ 0.10 g/l (10 mg/dl). The LOD = LOB + (c ∗ SD), where c = 1.645 / (1 − (1 / (4 ∗ df))) [5]. Leftover, de-identified serum samples with assigned triglyceride concentrations were obtained from a local laboratory and used for validation studies. To optimize sample/reagent volumes, 250 μl of a serum pool with an apoB concentration of ~200 mg/dl was added to each of six microtubes. Antisera reagent (25, 50, 100, 150, 200, or 250 μl) was added to the microtubes. Samples were vortexed and incubated for 10 min at room temperature, and then microcentrifuged for 10 min at 12,000 rpm (Eppendorf Model 5417C). The supernatants were transferred to Cobas Fara tubes for the measurement of cholesterol (CHOL), apolipoprotein AI (apoAI), and apoB in duplicate. To determine the effect of incubation time 250 μl of a serum pool was added to each of 5 microtubes followed by 250 μl antisera reagent. Samples were vortexed and allowed to incubate for 1, 5, 10, 30, and 60 min at room temperature, and then microcentrifuged for 10 min at 12,000 rpm. Supernatants were transferred to Cobas Fara tubes for the measurement of CHOL, apoAI, and apoB in duplicate. To determine the effect of centrifugation time, 250 μl of a serum pool was added to each of 3 microtubes, followed by 250 μl of antisera reagent. Samples were vortexed and incubated for 10 to 15 min at room temperature, and then microcentrifuged for 5, 10, or 15 min at 12,000 rpm. Supernatants were transferred to Cobas Fara tubes for the measurement of cholesterol, apoAI, and apoB in duplicate. To determine the effect of centrifugation speed 250 μl of a serum pool was added to each of four microtubes, followed by an equal volume of antisera reagent. Samples were vortexed and incubated for 10 min at room temperature, then microcentrifuged for 10 min at 8000, 10,000, 12,000, or 14,000 rpm. Supernatants were transferred to Cobas Fara tubes for the measurement of CHOL, apoAI, and apoB in duplicate. After optimization, subsequent studies were performed with the following procedure. First, 200 μl of the IP reagent was added to 200 μl of sample in a microcentrifuge tube and vortexed. After 10 min incubation at room temperature the tube was microcentrifuged at 12,000 rpm for 10 min. The supernatants were transferred to Cobas Fara sample tubes for the measurement of apoAI, apoB, and CHOL. To assess the capacity of IP for apoB precipitation a serum based calibrator (Midland Bioproducts) with a very high total apoB (~ 300 mg/dl) was tested as described above. HDL cholesterol results for 118 serum samples were compared after IP and DS/MgCl2 precipitation. Because precipitation removed apoBcontaining lipoproteins, assaying total cholesterol in the supernatant provides a direct measurement of HDL cholesterol. Linear regression was used to compare methods. Robustness of IP and chemical precipitation was determined with 52 high triglyceride specimens (triglyceride concentrations from 2.86 to 23.63 mmol/l, 253 to 2091 mg/dl) to compare the frequency of incomplete precipitation. A solid pellet with a clear supernatant after precipitation was considered a successful separation. In some samples, some precipitate did not form a solid pellet, but it was still possible to remove the clear supernatant. If precipitation was unsuccessful, additional IP or DS/MgCl2 was added or fresh sample was diluted in half before repeating the precipitation procedure. Dextran sulfate/magnesium chloride precipitation was performed as described by Kimberly [6]; 30 μl reagent was added to 300 μl of a sample, vortexed, incubated for 10 min at room temperature, and then microcentrifuged at 10,000 rpm for 5 min.

349

Fig. 1. Immunoprecipitation of apoB-containing lipoproteins in a 250 μl serum pool with various volumes of antiserum reagent. An equal volume of the reagent completely precipitates all apoB-containing lipoproteins. The serum pool contained 3.02 g/l (302 mg/dl) apoB, 2.79 g/l (279 mg/dl) apoAI, and 16.3 mmol/l (630 mg/dl) total cholesterol.

3. Results and discussion Our apoB immunoturbidimetric assay had an LOB and LOD of 0.018 g/l (1.8 mg/dl) and 0.111 g/l (11.1 mg/dl), respectively. Dose response studies indicated that equal volumes of a sample and reagent completely precipitated all apoB-containing lipoproteins with no effect on HDL (Fig. 1). Equal volumes of a sample and the IP reagent precipitated all apoB lipoproteins in the serum-based calibrator (3.02 g/l apoB) without any coprecipitation of apoAI. The mean recovery of apos AI and B were 102.5% and 3.7%, respectively. There was little effect of incubation time, centrifugation speed, or centrifugation time on immunoprecipitation using antiserum (Table 1). The mean recovery of apos AI and B after IP was 95.4% and 0.7%, respectively; all apoB results were less than the limit of detection of the assay. ApoAI recovery was essentially 100%. Where DS/MgCl2 precipitation was successful, agreement by IP was very good (Fig. 2). However, IP was much more robust in isolating HDL particles in high triglyceride samples (Table 2). Whereas IP successfully removed all apoB-containing lipoproteins in samples with triglycerides up to about 22.6 mmol/l (2000 mg/dl), DS/MgCl2 was successful only

Table 1 Effect of incubation time, centrifugation speed, and centrifugation time on immunoprecipitation. CHOL (mmol/l)

ApoAI (g/l)

ApoB (g/l)

% ApoB remaining

Incubation time (min) Sample 9.71 1 2.96 5 2.84 10 2.85 30 2.83 60 2.81

3.10 3.53 3.69 3.92 3.81 3.65

2.01 0.05 0.04 0.04 0.04 0.03

2.3% 2.0% 1.7% 1.7% 1.6%

Centrifugation speed (rpm) Sample 10.18 8000 3.01 10,000 2.98 12,000 3.12 14,000 2.95

3.11 3.76 3.69 3.44 3.38

1.99 0.00 0.01 0.00 0.00

0.1% 0.6% 0.0% 0.1%

Centrifugation time (min) Sample 10.18 5 2.62 10 2.74 15 2.79

3.32 3.46 3.38 3.49

2.03 0.03 0.04 0.04

1.5% 1.9% 1.8%

350

J.H. Contois et al. / Clinica Chimica Acta 436 (2014) 348–350 Table 2 Comparison of successful HDL separation; immunoprecipitation vs. dextran sulfate/ magnesium chloride precipitation of samples with high triglyceride concentrations. TG, mmol/l

N

IP 1st

IP 2nd

DS 1st

DS 2nd

2.83–4.51 (250–399 mg/dl) 4.52–7.90 (400–699 mg/dl) 7.91–11.26 (700–999 mg/dl) N11.30 (1000 mg/dl) ALL

11 9 14 18 52

100.0% 77.8% 85.7% 22.2% 65.4%

– 100% 100% 61.1% 86.5%

18.2% 0.0% 0.0% 5.6% 5.8%

100% 55.6% 14.3% 16.7% 40.4%

Key: 1st; 1st attempt with usual procedure, 2nd; 2nd attempt with additional reagent and/ or sample dilution prior to precipitation, IP; immunoprecipitation, DS: dextran sulfate/ magnesium chloride reagent, TG; triglycerides.

Fig. 2. Comparison of immunoprecipitation (Y) and dextran sulfate/magnesium chloride precipitation (X) for HDL cholesterol measurement. Regression equation: y = 1.01x + 0.07, r = 0.99.

up to about 5.6 mmol/l (500 mg/dl) triglycerides, and even then only after repeat testing with additional reagent or sample dilution. Chemical precipitation methods are generally effective in separating HDL particles from apoB-containing lipoproteins. Typically, polyanions such as heparin, dextran sulfate, and sodium phosphotungstate are used with a divalent cation, such as magnesium or manganese. The DS/MgCl2 precipitation method, using dextran sulfate with an approximate molecular weight of 50,000 Da, is the most popular precipitation method in the US, and is recognized as a component of the designated comparison method for HDL cholesterol measurement [6]. Unfortunately, with chemical methods, triglyceride-rich lipoproteins may not completely precipitate, as evidenced by turbidity in the supernatant. According to Warnick and colleagues, chemical precipitation methods may slightly overestimate HDL cholesterol due to incomplete precipitation of VLDL and LDL [7]. High triglyceride samples are problematic presumably because the higher densities of the TRLs make sedimentation by centrifugation difficult. In one study, the percentage of samples requiring additional treatment because of incomplete precipitation was 4%, 7.5%, 10%, 11%, and 12% for PEG (10%), DS/MgCl2, heparin/ manganese, and PEG (7.5%), respectively [7]. Warnick et al. determined that both DS/MgCl2 and heparin/manganese left small amounts of apoB in the supernate, while precipitating small amounts of HDL [8]. Dextran sulfate precipitated slightly more apoB-containing lipoproteins than heparin, but also precipitated slightly more HDL [8]. HDLs are heterogeneous particles that vary in size and composition, without a definitive chemical structure, making development of specific assays for cholesterol measurement difficult. In the most comprehensive study to date, Miller et al. compared direct methods for LDL cholesterol and HDL cholesterol with the “beta-quant” reference method using fresh samples from subjects with and without cardiovascular disease and/or various lipid and lipoprotein disorders [4]. These data showed a dramatic lack of agreement between methods and in comparison with the reference method, especially with specimens from patients with dyslipidemias and cardiovascular disease, the so called “diseased” specimens. Six of eight direct HDL cholesterol assays failed to meet NCEP total error goals in the healthy control group, while all eight assays failed to meet performance goals in the group with cardiovascular disease and/or lipoprotein disorders [4]. Total variability ranged from 2.6% to 16.4%, total error ranged from − 8.2% to 36.3%, and mean bias ranged from − 8.6% to 8.8% between assays and the

reference method [4]. Chemical precipitation and especially immunoprecipitation offer advantages over direct methods in terms of accuracy. Monospecific antibodies directed against apolipoproteins are the most specific method available for the separation of lipoproteins. Puchols et al. compared ultracentrifugation and chemical precipitation methods to immunoprecipitation for the isolation of HDL and reported that only immunoprecipitation completely separated apoAI and apoB containing particles [9]. Chemical precipitation and ultracentrifugation failed to separate 4% to 20% of HDL particles [9]. Similar results were reported by Heuck et al. with immunoprecipitation: no beta or prebeta lipoproteins were present in the supernatant after immunoprecipitation, with only alpha lipoproteins remaining [10]. Our data are consistent with these previous studies showing that immunoprecipitation is specific in isolating HDL particles, and also prove that immunoprecipitation is more robust in separating triglyceride-rich lipoproteins. Because of the specificity of anti-apoB antibodies, HDL particles will not co-precipitate with apoB-containing lipoproteins, which may be an issue with chemical precipitation methods. The availability of a validated, commercially available immunoprecipitation reagent should prove useful to clinicians and researchers with a need to accurately separate HDL particles.

References [1] Movva R, Rader DJ. Laboratory assessment of HDL heterogeneity and function. Clin Chem 2008;54:788–800. [2] Vaisar T, Pennathur S, Green PS, Gharib SA, Hoofnagle AN, Cheung MC, et al. Shotgun proteomics implicates protease inhibition and complement activation in the antiinflammatory properties of HDL. J Clin Invest 2007;117:746–56. [3] Warnick GR, Nauck M, Rifai N. Evolution of methods for measurement of HDLcholesterol: from ultracentrifugation to homogeneous assays. Clin Chem 2001;47:1579–96. [4] Miller WG, Myers GL, Sakurabayashi I, Bachman LM, Caudill SP, Dziekonski A, et al. Seven direct methods for measuring HDL and LDL cholesterol compared with ultracentrifugation reference measurement procedures. Clin Chem 2010;56:977–86. [5] NCCLS. Protocols for determination of limits of detection and limits of quantitation; approved guideline. NCCLS document EP17-A. 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898 USA: NCCLS1-56238-551-8; 2004. [6] Kimberly M, Leary E, Cole T, Waymack P. Selection, validation, standardization, and performance of a designated comparison method for HDL cholesterol for use in the cholesterol reference method laboratory network. Clin Chem 1999;45:1803–12. [7] Warnick GR, Nguyen T, Albers AA. Comparison of improved precipitation methods for quantification of high-density lipoprotein cholesterol. Clin Chem 1985;31:217–22. [8] Warnick GR, Benderson J, Albers JJ. Dextran sulfate-Mg2+ precipitation procedure for quantitation of high-density-lipoprotein cholesterol. Clin Chem 1982;28:1379–88. [9] Puchols P, Luley C, Alaupovic P. Comparison of four procedures for separating apolipoprotein A- and apolipoprotein B-containing lipoproteins in plasma. Clin Chem 1987;33:1597–602. [10] Heuck C-C, Erbe I, Mathias D. Cholesterol determination in serum after fractionation of lipoproteins by immunoprecipitation. Clin Chem 1985;31:252–6.