Measurement of LDL particle size in whole plasma and serum by high performance gel-filtration chromatography using a fluorescent lipid probe

Departments of
¹ Clinical Chemistry and
² Endocrinology, Research Institute for Endocrinology, Reproduction and Metabolism, Academic Hospital Vrije Universiteit, P. O. Box 7057, 1007 MB Amsterdam, The Netherlands.
^a Author for correspondence. Fax 31-20-4443895;

	Abstract

Top Abstract Introduction Materials and Methods Results and Discussion References

 
We have recently described a technique for measuring LDL size by high performance gel-filtration chromatography (HPGC) with UV detection (Scheffer et al., Clin Chem 1997;43:1904–12). A drawback of this method is the necessity of LDL isolation before chromatography. We now describe a modification of this method based on selective detection of lipoproteins by postcolumn labeling with parinaric acid, a fluorescent lipid probe. Measuring the size of isolated LDL by HPGC in 56 subjects, we obtained diameters of 25.72 ± 0.60 nm with UV detection and of 25.74 ± 0.58 nm with fluorescence detection. The modified method is suitable for LDL size measurement in whole plasma or serum. LDL sizes measured in whole plasma correlated strongly with the respective values in isolated LDL (r = 0.976) but were on average 0.18 nm larger (P < 0.001). CVs for within- and between-series imprecision were <0.25%. The present method requires only 5 µL plasma or serum without sample preparation and is suitable for the unattended analysis of large series of samples.

	Introduction

Top Abstract Introduction Materials and Methods Results and Discussion References

 
LDLs have been shown to be a heterogeneous population of particles with respect to size, density, and composition. The results from several case-control studies have demonstrated that small, dense LDL is associated with increased risk of developing coronary heart disease (1)(2)(3)(4) and non-insulin-dependent diabetes mellitus (5)(6). A variety of methods is available for measurement of LDL particle size. The most commonly used technique is non-denaturing gradient gel electrophoresis (GGE).¹ GGE can be performed on isolated LDL or on whole plasma (7)(8)(9). However, GGE is a labor-intensive method that is less suitable for the analysis of a large series of samples. We have recently described an alternative technique for measuring the mean LDL particle diameter by high performance gel-filtration chromatography with detection by UV absorbance at 280 nm (HPGC-UV) (10). The HPGC-UV method is very reproducible, precise, and suitable for analyzing a large series of samples. Mean particle diameters measured by HPGC-UV were in close agreement with peak particle diameter values obtained by GGE. Because detection is based on absorbance at 280 nm, the method requires LDL isolation by ultracentrifugation before chromatography to avoid interference by plasma proteins.

This study reports on a modification of the HPGC-UV technique to make it suitable for the measurement of LDL size in whole plasma and serum, by postcolumn labeling of lipoproteins with the specific fluorescent lipid probe cis-parinaric acid (PnA). PnA is a fatty acid with four conjugated double bonds, showing intense fluorescence in a lipid environment. PnA has been shown to be spontaneously incorporated into the lipid matrix of lipoproteins, resulting in a dramatic increase in fluorescence intensity (11). PnA incorporation into LDL particles has been used by other investigators to study kinetics of LDL oxidation (12). The PnA reagent is particularly suited for postcolumn detection of lipoproteins after separation by HPGC, because the fluorescence intensity of the probe in aqueous solution is very low and is enhanced manyfold on incorporation into lipoproteins.

In the present study, we report on the results of a comparison of the original HPGC-UV method with HPGC in combination with fluorescence detection (HPGC-FL).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results and Discussion References

 
plasma samples
Blood was obtained from 24 healthy medical students and 32 non-insulin-dependent diabetes mellitus patients with acceptable glycemic control (HbA1c, 6.2% ± 1.1%). Venous blood samples were collected into Vacutainer Tubes containing EDTA (final concentration, ~4 mmol/L). For the comparison between plasma and serum samples, additional blood samples were obtained from 20 medical students. After low-speed centrifugation at room temperature, serum and plasma samples were stored at -86°C until analysis. The study was approved by our hospital ethics committee, and all subjects gave informed consent to take part in the study.

chemicals
PnA (9,11,13,15-cis-trans-trans-cis-octadecatetraenoic acid) was obtained from Molecular Probes. Stock solutions of PnA dissolved in ethanol and stored at -20°C could be used for several months. Working solutions of PnA were freshly prepared in degassed TBE buffer (90 mmol/L Tris, 80 mmol/L boric acid, and 3 mmol/L EDTA, adjusted to pH 9.6). All reagents, buffer components, and HPLC solvents were analytical grade.

ldl isolation
LDL (d, 1.019–1.063 kg/L) isolation was performed by sequential preparative ultracentrifugation in a Beckman Optima TLX ultracentrifuge with a type 100.4 rotor. In the first step of LDL isolation, plasma was adjusted to d = 1.019 kg/L. VLDL and intermediate density lipoprotein were removed by aspiration after the first ultracentrifugation run (267 000g_av, 2 h 40 min, 15 °C). The infranate was transferred to a new tube and adjusted to d = 1.063 kg/L. The tubes were centrifuged for a second run (176 000g_av, 16 h, 15 °C). Afterwards, the LDL fraction was collected from the top by aspiration, stored at 4°C, and analyzed within 4 days. A DMA 38 density meter from Paar Physica was used to check the density of the solutions.

hpgc
The HPLC system was composed of a Model 616 pump, Model 486 UV detector, Model 747 fluorescence detector, a Model 717 autosampler from Waters, and a Degasys Model DG2410 mobile phase degasser from Uniflows. Millennium 2010 software from Waters was used for instrument control and data acquisition and processing. The sample compartment of the autosampler was cooled at 7 °C. Chromatography was performed using a Superose 6 HR 10/30 column from Pharmacia eluted with phosphate-buffered saline (0.1 mol/L NaH₂PO₄·H₂O, 0.2 mol/L NaCL, and 0.1 mmol/L disodium EDTA; adjusted to pH 7.4) at a flow rate of 0.5 mL/min. The column was kept at a constant temperature of 25°C in a water bath. The retention time of the LDL peak was used to calculate the mean LDL particle diameter. One control and two calibration LDL samples stored in aliquots at -86°C were included in every series of samples.

HPGC-UV measurements were carried out using LDL isolated by ultracentrifugation. Every 50 min a sample (50 µL) was injected. The column effluent was monitored by UV detection at 280 nm. The entire procedure was recently described in detail (10).

The HPGC-FL technique is a modification of the HPGC-UV method. Instead of isolated LDL, whole plasma or serum (5 µL) was injected on the Superose column. PnA reagent, at a flow rate of 0.05 mL/min, was continuously added to the column effluent via a low dead-volume mixing-tee (final PnA, concentration 1 µmol/L). To provide a pulse-free, reproducible flow, PnA addition was carried out using a syringe pump (Model STC-521, Terumo). Between the mixing-tee and the fluorescence detector, a 0.5-mL sample loop was installed to allow for a sufficient reaction time (~1 min) for labeling of lipoproteins by PnA. Fluorescence was measured with excitation at 324 nm and emission at 413 nm. The total run time was 30 min.

statistics
All data were analyzed using SPSS statistical software. Results are presented as mean ± SD unless specified otherwise. Pearson's correlation coefficients were computed, and linear regression analyses were performed to assess associations. Methods were compared using paired Student's t-test. The nonparametric paired Wilcoxon rank test was used to compare serum and plasma LDL particle diameters. P <0.05 was considered significant.

	Results and Discussion

Top Abstract Introduction Materials and Methods Results and Discussion References

 
hpgc-uv vs hpgc-fl
We reported previously (10) that LDL size measured by HPGC-UV is in good agreement with values obtained by GGE, which is considered the conventional assay system for measuring LDL size. Interference by coeluting plasma proteins detected at 280 nm precludes the use of the HPGC-UV technique for the measurement of LDL size in whole plasma or serum. In the present study, we have attempted to overcome this limitation by using more selective detection, i.e., postcolumn labeling with PnA. Precolumn labeling of lipoproteins was not considered a viable alternative, because incorporation of a chromophoric or fluorescent probe into the LDL particle may lead to an increase in size. As illustrated in Fig. 1 , the main lipoprotein classes VLDL, LDL, and HDL in plasma are well separated after postcolumn labeling of lipoproteins with PnA in combination with fluorescence detection. Control experiments with lipoprotein-free plasma revealed that albumin seemed to be one of the few plasma proteins with binding affinity for PnA (Fig. 1 , broken line). Most importantly, in the region of the chromatogram where LDL elutes (18–25 min), no PnA-binding plasma proteins were detected. An additional advantage of fluorescence detection after PnA labeling is that the time between injections could be reduced to 30 min compared with 60 min for the original HPGC-UV procedure (10).

View larger version (17K): [in this window] [in a new window]   Figure 1. High performance gel-filtration chromatography. Shown is a profile of whole plasma (solid line) and lipoprotein-free plasma (broken line) prepared by ultracentrifugation at d >1.23 kg/L, with fluorescence detection (excitation, 324 nm; emission, 413 nm), after postcolumn labeling with PnA (HPGC-FL).

We used plasma from 56 subjects to evaluate the HPGC-FL method further. Non-insulin-dependent diabetes mellitus patients were included to create a wide range of LDL sizes, because diabetic individuals have smaller LDL particles than nondiabetic individuals (5)(6). LDL was isolated from these samples, and LDL size was determined by both HPGC-UV and HPGC-FL methods (Table 1 ). The 0.02-nm average difference observed between both methods was not statistically significant (P = 0.294). Values obtained by HPGC-FL were in close agreement with values obtained by HPGC-UV [y = (0.93 ± 0.03)x (1.79 ± 0.86); r = 0.967; S_yx = 0.15 nm]. In addition, LDL size was determined using direct injection of whole plasma by HPGC-FL (Table 1 ). As shown in Fig. 2 A, measurements by HPGC-FL using whole plasma were closely correlated to measurements by HPGC-UV using isolated LDL [y = (0.97 ± 0.04)x (1.00 ± 0.98); r = 0.961; S_yx = 0.17 nm]. However, we obtained statistically greater values measuring the LDL size in whole plasma than in isolated LDL. The difference between the measurements by the two methods was not related to the actual LDL size, as illustrated in Fig. 2B (13). It was also excluded that this difference was caused by a matrix effect of plasma on the chromatographic properties of LDL, because addition of lipoprotein-free plasma to isolated LDL did not alter the retention time of the LDL peak (data not shown). In good agreement with our study, Westhuyzen et al. (14) found slightly larger LDL peak particle diameters using plasma than corresponding values derived from isolated LDL, as measured by GGE. The difference between LDL size measured in plasma and in LDL after isolation by ultracentrifugation is possibly due to effects of the gravitational field at high rotor speeds on lipoprotein structure. Recently, two studies (15)(16) reported that the content of triglycerides, phospholipids, and especially cholesterol of LDL tended to decrease with increasing rotor speed. Taken together, these findings suggest that ultracentrifugation of plasma lipoproteins may produce a reduction of LDL size.

View larger version (17K): [in this window] [in a new window]   Figure 2. HPGC-UV and HPGC-FL measurements. (A) Plot showing the correlation of LDL size measurements by HPGC using whole plasma and fluorescence detection after labeling with PnA (HPGC-FL) vs measurements using isolated LDL with UV detection (HPGC-UV) with its regression line (solid line) and the line of unity (broken line). (B) Bias plot of absolute differences (HPGC-UV minus HPGC-FL) vs the average of the two methods, with the mean of the differences (solid line) and mean ± 2SD limits (broken line).

serum vs plasma
To study the influence of blood clotting on LDL size measurements, we performed a study among 20 healthy subjects. Plasma and serum were collected, and the mean LDL size of these subjects was measured by HPGC-FL. There was no significant difference (P = 0.586) in LDL size between plasma and serum. The mean LDL size measured in plasma was 26.36 ± 0.27 nm, almost equal to the value of 26.37 ± 0.29 nm measured in serum [y = (1.03 ± 0.07)x - (0.72 ± 1.73); r = 0.965; S_yx = 0.08 nm].

reproducibility
Precision of the HPGC-FL method was determined by calculating both within-run and between-run CVs (n = 10). For calculation of the within-run reproducibility, fresh samples were used. By using isolated LDL and whole plasma, we found that the CVs were 0.14% and 0.22%, respectively. The between-run CV calculated from measurements performed on different days, using isolated LDL samples stored in aliquots at -86°C, was 0.21%.

probe conditions
At low PnA concentrations, there is a linear relation between the final concentration of PnA and the fluorescence intensity of the PnA-labeled LDL peak, which deviates at ~2 µmol/L because of increasing self-quenching of the probe in the lipid environment. We used final PnA concentrations of 1, 3, and 10 µmol/L, and almost equal sizes were measured of an isolated LDL sample (25.63, 25.60, and 25.64 nm, respectively), demonstrating that the PnA concentration is not a critical parameter. We routinely used a final PnA concentration of 1 µmol/L, which gave a proper fluorescence signal. The postcolumn PnA reagent was prepared by diluting an ethanolic PnA stock solution in TBE buffer adjusted to pH 9.6. At this pH, the PnA molecule is ionized and has a higher solubility in water. Under these conditions, during an analysis series of 16 h, maintenance of a stable signal of the fluorescent lipoprotein-PnA complex was achieved. In contrast, when PnA was diluted in phosphate-buffered saline adjusted to pH 7.4, the fluorescence intensity of the PnA-labeled LDL peak decreased ~40% over an 8-h period, probably by nonspecific adsorption of the probe to the reagent reservoir or tubing.

The main advantage of the HPGC-FL method is that it permits direct measurement of LDL size in whole plasma or serum. The procedure is suitable for very precise and reproducible measurement of LDL size from a very small amount of sample (5 µL). No sample pretreatment steps are involved in the procedure. Because analysis of a single sample takes only 30 min, a large series of samples can be analyzed using standard HPLC equipment with minimal hands-on time.

	Acknowledgments

 
We thank J.C.T. Meeues for his cooperation in collecting blood samples of healthy volunteers and E.E. Musch for isolation of LDL by ultracentrifugation.

	Footnotes

 
¹ Nonstandard abbreviations: GGE, gradient gel electrophoresis; HPGC, high performance gel-filtration chromatography; PnA, parinaric acid; and FL, fluorescence.

	References

Top Abstract Introduction Materials and Methods Results and Discussion References

 

1. Tornvall P, Karpe F, Carlson LA, Hamsten A. Relationships of low density lipoprotein subfractions to angiographically defined coronary artery disease in young survivors of myocardial infarction. Atherosclerosis 1991;90:67-80. [ISI][Medline] [Order article via Infotrieve]
2. Gardner CD, Fortmann SP, Krauss RM. Association of small low-density lipoprotein particles with the incidence of coronary artery disease in men and women. JAMA 1996;276:875-881. [Abstract]
3. Stampfer MJ, Krauss RM, Ma J, Blanche PJ, Holl LG, Sacks FM, et al. A prospective study of triglyceride level, low-density lipoprotein particle diameter, and risk of myocardial infarction. JAMA 1996;276:882-888. [Abstract]
4. Lamarche B, Tchernof A, Moorjani S, Cantin B, Dagenais GR, Lupien PJ, et al. Small, dense low-density lipoprotein particles as a predictor of the risk of ischemic heart disease in men. Prospective results from the Quebec Cardiovascular Study. Circulation 1997;95:69-75. [Abstract/Free Full Text]
5. Singh ATK, Rainwater DL, Haffner SM, Van deBerg JL, Shelledy WR, Moore PH, Jr, et al. Effect of diabetes on lipoprotein size. Arterioscler Thromb Vasc Biol 1995;15:1805-1811. [Abstract/Free Full Text]
6. Austin MA, Mykkänen L, Kuusisto J, Edwards KL, Nelson C, Haffner SM, et al. Prospective study of small LDLs as a risk factor for non-insulin dependent diabetes mellitus in elderly men and women. Circulation 1995;92:1770-1778. [Abstract/Free Full Text]
7. Krauss RM, Burke DJ. Identification of multiple subclasses of plasma low density lipoproteins in normal humans. J Lipid Res 1982;23:97-104. [Abstract]
8. Rainwater DL, Moore PH, Jr, Shelledy WR, Dyer TD, Slifer SH. Characterization of a composite gradient gel for the electrophoretic separation of lipoproteins. J Lipid Res 1997;38:1261-1266. [Abstract]
9. Nichols AV, Krauss RM, Musliner TA. Nondenaturing polyacrylamide gradient gel electrophoresis. Methods Enzymol 1986;128:417-431. [ISI][Medline] [Order article via Infotrieve]
10. Scheffer PG, Bakker SJ, Heine RJ, Teerlink T. Measurement of low-density lipoprotein particle size by high-performance gel-filtration chromatography. Clin Chem 1997;43:1904-1912. [Abstract/Free Full Text]
11. Van den Berg JJM. Effects of oxidants and antioxidants evaluated using parinaric acid as a sensitive probe for oxidative stress. Redox Rep 1994;1:11-21.
12. Tribble DL, Krauss RM, Lansberg MG, Thiel PM, Van den Berg JJM. Greater oxidative susceptibility of the surface monolayer in small dense LDL may contribute to differences in copper-induced oxidation among LDL density subfractions. J Lipid Res 1995;36:662-671. [Abstract]
13. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;i:307-310.
14. Westhuyzen J, Graham SD, Rasiah RL, Saltissi D. Simplified sizing of low-density lipoprotein using polyacrylamide gradient gel electrophoresis of plasma. Eur J Clin Chem Clin Biochem 1997;35:17-19. [Medline] [Order article via Infotrieve]
15. Leonhardt W, Pietzsch J, Nitzsche S. Very-fast ultracentrifugation of human plasma lipoproteins: influence of the centrifugal field on lipoprotein composition. Clin Chim Acta 1994;224:21-32. [ISI][Medline] [Order article via Infotrieve]
16. Pietzsch J, Subat S, Nitzsche S, Leonhardt W, Schentke KU, Hanefeld M. Very fast ultracentrifugation of serum lipoproteins: influence on lipoprotein separation and composition. Biochim Biophys Acta 1995;1254:77-88. [Medline] [Order article via Infotrieve]