E L S E V I E R Clinica Chimica Acta 258 (1997) 219-229
Effect of storage at - 70°C on lipid, l ipoprotein and apolipoprotein concentrations
Kevin Evans*, Janice Mitcheson, Michael F. Laker Department of Clinical Biochemistry and Metabolic Medicine, University of Newcastle upon Tyne,
Newcastle upon Tyne, UK
Received 10 May 1996; revised 3 October 1996; accepted 16 October 1996
Abstract
We have investigated the effects on lipid, apolipoprotein and lipoprotein measurements of storing unfractionated serum from normolipidaemic and hyperlipidaemic subjects at - 70°C for 10 days., 3 months and 6 months. Total serum concentrations of lipids and apolipo- proteins were stable except for triglyceride concentrations. These increased on storage although the change was < 2.0% after 6 months. Storage of serum before sequential flotation ultracentrifugation resulted in decreased free cholesterol and phospholipid concentrations in very low density lipoproteins. In low density lipoproteins, free cholesterol concentrations increased and protein concentrations decreased on storage, while esterified cholesterol and phospholipi,d concentrations fell after 10 days but did not differ from baseline concentrations after storag;e for 6 months. Within high density lipoproteins, there were decreases in triglyceride and protein concentrations. Although storage of serum at - 70°C for up to 6 months did not result in extensive changes in most lipoprotein fractions, separation of lipoprotein fractions from serum should, ideally, be performed as soon as possible after collection. O 1997 Elsevier Science B,V. All rights reserved
Keywords: Cholesterol; Triglycerides; Phospholipids; Sample handling; Variation, source of
* Corresponding author. Department of Clinical Biochemistry, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. Tel: +44 1865 221109; fax: +44 1865 220348.
220 K. Evans et al. / Clinica Chimica Acta 258 (1997) 219 229
1. Introduction
Serum total cholesterol and triglyceride concentrations have been shown to be stable on storage at - 20°C for up to 5 years [1]. However, fewer data are available on the effects of stor~/ge of unfractionated serum on lipo- protein cholesterol and triglyceride concentrations. Much of the work in this field has been concerned with high-density lipoprotein (HDL) choles- terol concentrations determined after precipitation of apolipoprotein (apo) B-containing lipoproteins in stored serum [2-7]. Relatively few studies have investigated the effect of serum storage on lipoprotein composition deter- mined by ultracentrifugation. Ideally, the procedure should be performed on fresh serum, although the number of specimens obtained from dynamic function tests or large studies may exceed the capacity for immediate analysis; thus, storage of unfractionated serum would be desirable. Studies to date have been contradictory, with reports that lipoproteins are stable if serum is stored at 0-5°C for 2 weeks but show changes after only a few days at - 2 8 ° C [8]. Others have suggested that cholesterol concentrations are stable for at least 27 weeks at - 2 0 ° C and triglyceride concentrations are stable for at least 11 weeks, although the subjects studied had relatively normal total serum lipid concentrations [9].
In a previous report we showed that extensive changes occurred in lipoprotein fractions after storing serum for 10 days at 4°C or 3 months at - 2 0 ° C before lipoprotein separation by sequential flotation ultracentrifu- gation [10]. We have, therefore, extended this work to investigate the effect of storing serum at - 70°C from subjects with a wide range of total serum lipid concentrations on lipid, lipoprotein and apolipoprotein concentrations.
2. Subjects and methods
2.1. Clinical protocol
The study was approved by the Joint Ethics Committee of the University of Newcastle and Newcastle Health Authority, and all subjects gave in- formed consent verbally. Forty-seven subjects (22 men and 25 women) with a wide range of total serum lipid concentrations were recruited from the Metabolic Clinic, Royal Victoria Infirmary, Newcastle upon Tyne, and from University staff. Mean total serum lipid concentrations were choles- terol 7.11 mmol/1 (range 4.52-12.02) and triglycerides 2.08 mmol/1 (range 0.58-7.41). Fourteen subjects were normolipidaemic (cholesterol <6.5 mmol/1, triglycerides < 1.8 mmol/1); ten subjects were hypercholestero- laemic (cholesterol > 6.5 mmol/1, triglycerides < 1.8 mmol/1); 13 subjects
K. Evans et al. / Clinica Chimica Acta 258 (1997) 219-229 221
Table 1 Changes in serum concentrations of total cholesterol, triglycerides, apo A-I and apo B on storage at - 70°C
1) Apo A-I (g/l) 1.57 (0.24) 1.52 (0.19) 1.53 (0.26) 1.54 (0.26) 0.21 Apo B (g/l) 1.31 (0.44) 1.29 (0.45) 1.27 (0.43) 1.27 (0.43) 0.22
Values are mean (S.D.), 47 subjects were studied. Statistical analysis: ANOVA with repeated measures design.
had combined hyperlipidaemia (cholesterol > 6.5 mmol/1, triglycerides > 1.8 mmol/1) and ten subjects had hypertriglyceridaemia (cholesterol < 6.5 mmol/1, triglycerides > 1.8 mmol/1).
Blood (40 ml) was collected with minimal venous stasis from an antecu- bital vein after a 12-h overnight fast. Blood was allowed to clot, and serum separated by centrifugation at 800 x g for 15 min. Serum samples were then divided into 3-ml aliquots and preserved with a 'cocktail' of 1 mmol/1 EDTA (BDH Laboratory Supplies, Lutterworth, Leics, UK) to chelate
Table 2 Changes in lipid and protein concentrations in VLDL on storage at - 70°C
Values are mean (S.D.), 47 subjects were studied. Statistical analysis: ANOVA with repeated measures design.
heavy metals [11], 1 mmol/1 phenylmethylsulfonyl fluoride (BDH) in di- methylsulfoxide (BDH) as an inhibitor of serine proteases [12] and 0.1 mmol/1 butylated hydroxytoluene (Sigma, Poole, Dorset, UK) in absolute ethanol (Hayman Limited, Witham, Essex, UK) as a lipid antioxidant [13]. These preservatives were added to minimise changes in lipoproteins occurring during ultracentrifugation.
2.2. Laboratory method
Serum was separated into lipoprotein fractions by sequential flotation ultracentrifugation in a Beckman 70.1Ti fixed angle rotor at 20°C and 47000 rev./min (100000xg) [12]. Four 24-h steps were necessary to separate very low density lipoproteins (VLDL, d < 1.006 kg/1), intermediate density lipoproteins (IDL, 1.006-1.019 kg/1), low density lipoproteins (LDL, 1.019-1.063 kg/1) and HDL (1.063-1.21 kg/1). Eight ml of 0.189 mol/1 sodium chloride (Sigma) was added to 2 ml of serum to give a density of 1.006 kg/1. After ultracentrifugation VLDL was removed in the top 2 ml. The density was then adjusted to 1.019 kg/1 with sodium chloride to separate IDL, after' IDL removal the density was adjusted to 1.063 kg/1 with sodium chloride to separate LDL. To isolate HDL, the density was adjusted to 1.21 kg/1 with sodium bromide (Sigma). Ultracentrifugation of one serum aliquot was begun within 3 days; other aliquots were stored at - 7 0 ° C for 10 days, 3 months and 6 months prior to ultracentrifugation and analysis.
K. Evans et al. / Clinica Chimica Acta 258 (1997) 219-229 223
Concentrations of total and free cholesterol, triglyceride, phospholipid, and protein were measured using a Cobas Bio centrifugal analyser (Roche, Welwyn Garden City, UK). Serum and lipoprotein cholesterol concentra- tions were measured using a cholesterol oxidase method (Boehringer Mannheim, Lewes, East Sussex, UK). Interassay CVs were 2.2% at 3.4 mmol/1, 1.9% at 7.9 mmol/1 and 2.0% at 8.8 mmol/1. Triglyceride concentra- tions in serum and lipoprotein fractions were measured using a lipase-glyc- erol kinase method (Roche). Interassay CVs were 10.4% at 0.12 mmol/1, 2.5% at 2.0 mmol/1 and 2.5% at 3.0 mmol/1. Free cholesterol and phospho- lipid concentrations in lipoprotein fractions were measured using enzymatic colorimetric methods (Boehringer Mannheim) [14,15]. The assays were optimised by adjusting reagent volumes to improve sensitivity. Interassay CV for free cholesterol was 5.1% at 0.87 mmol/1 and for phospholipids was 3.2% at 0.57 mmol/1. Protein concentrations in lipoprotein fractions were determined by a modification of the Lowry technique [16]. Interassay CV was 4.7% at 0.8 g/1. Apo A-I and apo B concentrations in serum were measured by rate nephelometry using a Technicon DPA-1 analyser (Techni- con, Basingstoke, Hants, UK). Interassay CVs for apo A-I were 8.3% at 0.89 g/l, 6.5% at 1.40 g/1 and 4.8% at 2.12 g/1. For apo B the interassay CVs were 3.9% at 0.60 g/l, 5.7% at 0.86 g/1 and 5.4% at 1.15 g/1.
o a •
-
i - --. ;, "2° o o "~ o ~ ' ~ . . •
.., • (2 ~ .
& .~ a a ~ . . ~
4g
0.0 O 1.0 1.5 2.0
B u e i i n e IDI.-lriglyceride concentrat ions (mmol/L.)
Values are mean (S.D.), 47 subjects were studied. Statistical analysis: ANOVA with repeated measures design.
2.3. Statistical methods
Data were analysed using the Statistical Package for Social Sciences (SPSS for Windows). The distribution of data was determined using the Kolmogorov-Smirnov goodness of fit procedure [17]. After establishing that the distribution of all variables, including triglycerides, did not differ from normal, changes in concentrations with storage were assessed using analysis of variance (ANOVA) with a repeated measures design. Pearson correlation coefficients were calculated to explore the relationship between changes in lipoprotein lipid concentrations and initial levels.
3. Results
The mean changes in serum total cholesterol, triglycerides, apo A-I and apo B concentrations under all storage conditions were < 4.0% (Table 1). There were significant increases in triglyceride concentrations on storage although these were less than 2.0%. The mean overall recoveries of choles- terol and triglycerides after ultracentrifugation of fresh serum were 82.4% and 84.0% respectively.
K. Evans et al. / Clinica Chimica Acta 258 (1997) 219-229 225
3.1. VLDL
Decreases in free cholesterol and phospholipid concentrations of 13% and 10%, respectively, were seen after storage at - 7 0 ° C (Table 2).
3.2. IDL
No significant changes in mean concentrations of any analyte occurred in IDL (Table 3). However, trends in the changes in all IDL analytes, after both 3 raonths and 6 months, showed negative correlations with baseline values, the correlations being strongest with IDL-triglyceride concentrations (Fig. 1).
3.3. L D L
Esteritied cholesterol and phospholipid concentrations decreased after 10 days, returning to baseline after 6 months storage (Table 4). Free choles- terol concentrations increased and protein levels decreased on storage. Changes in free cholesterol and cholesterol ester levels after 3 months storage were dependent on baseline HDL composition (Figs. 2 and 3).
Fig. 2. Changes in L D L - and H D L - f r e e 'cholesterol af ter s torage o f s e rum for 3 m o n t h s at
- 70°C related to baseline H D L - f r e e cholesterol concent ra t ions . LDL-f ree choles terol ( ( 3 - -
O) , y = 35.6x - 13.2, r = 0.47, P < 0.001. H D L - f r e e choles terol ( O - - O ) , y = - 65.2x + 26.4,
r = - 0.48, P < 0.001.
226 K. Evans et al. / Clinica Chimica Acta 258 (1997) 219-229
60,
,~, 25 '
O C
O 8
.a 0 E
O OI C m
• 0 0
° o o 0 o- • • ooo'°
,,, • O O n .if'°" ..4~"
n 0 •
1 ! i i i
0.0 0.5 1.0 1 .S 2.0 2.$
B u o l l n o HDL-cholostorol ~ t o r ¢oncontralions (mmolrn.)
Fig. 3. Changes in LDL- and HDL-choles tero l ester after storage of serum for 3 m o n t h s at - 7 0 ° C related to baseline HDL-choles tero l ester concentra t ions . LDL-choles te ro l ester ( O - - O ) , y = 2 3 . 2 x - 27.6, r = 0 . 5 4 , P < 0 . 0 0 0 1 . HDL-choles tero l ester ( O - - O ) , y = - 20.7x + 21.6, r = - 0.43, P < 0.01.
3.4. H D L
Storage at - 7 0 ° C resulted in decreased HDL-triglyceride and protein concentrations (Table 5). After 3 months, changes in the free cholesterol and cholesterol ester content of the HDL fraction were correlated negatively with baseline HDL composition (Figs. 2 and 3). Altered HDL-protein concentrations after both 3 and 6 months were correlated negatively with baseline levels (Fig. 4).
4. Discussion
We have shown total serum cholesterol, apo A-I and apo B concentra- tions are stable for 6 months if stored at - 7 0 ° C and, although serum triglyceride levels increased, the changes were less than 2%. These findings are consistent with previous studies which have shown that serum lipids are stable if stored at - 2 0 ° C [1], and that apo A-I levels do not change with storage at - 20°C [10] and at - 70°C [18]. Our data for apo B differ from investigations which have reported apparent increases in serum concentra-
K. Evans et al. / Clinica Chimica Acta 258 (1997) 219-229 227
tions when determined by radioimmunoassay [19,20]. It is possible that apparent changes in apo B on storage are due to changes in lipoprotein structure causing altered exposure of epitopes [19]. Thus, such changes may be dependent on the analytical method and antibody used, and stability needs to be established for each analytical method and storage condition.
More extensive changes in lipoprotein composition occurred when serum was stored at - 70°C prior to fractionation, these affecting LDL and HDL particularly. Changes in the cholesterol contents of LDL and HDL after 3 months were related to baseline HDL composition, LDL-free and -esterified cholesterol concentrations correlating positively with baseline HDL-choles- terol, while HDL-free and -esterified cholesterol concentrations showed negative correlations (Figs. 2 and 3). Although these findings may be explained, at least partially, by redistribution of lipids between lipoprotein fractions, this cannot be the sole explanation since protein concentrations changed in the lipoprotein fractions. This suggests that physical disruption of lipoprotein particles occurred, affecting the HDL fraction particularly. The changes seen after 6 months were no greater than those noted after 3 months.. Although the changes were significant, they were less marked than those we have reported previously following storage at -20°C.
In conclusion, storage of serum at - 70°C prior to separation of lipo- protein fractions by ultracentrifugation results in fewer changes in lipo- protein composition than storage at 4°C or -20°C. However, lipoprotein separations should, ideally, be undertaken using fresh samples.
Table 5 Changes in lipid and protein concentrations in HDL on storage at - 70°C
[1] Kuchmak M, Taylor L, Olansky AS. Suitability of frozen and lyophilized sera for cholesterol and triglyceride determinations. Clin Chim Acta 1982;120:261-271.
[2] Reckless JPD, Betteridge D J, Wu P, Payne B, Galton DJ. Effect of storage on plasma lipoproteins (letter). Lancet 1977;ii:350.
[3] Stokes YM, Salmond CE, Carpenter LM, Welby TJ. Stability of total cholesterol, high-density lipoprotein cholesterol and triglycerides in frozen sera. Clin Chem 1986;32:995-999.
[5] Bachorik PS, Walker R, Brownell KD, Stunkard AJ, Kwiterovich PO. Determination of high density lipoprotein-cholesterol in stored human plasma. J Lipid Res 1980;21:608- 616.
[6] Bausserman LL, Saritelli AL, Milosavljevic D. High-density lipoprotein subfractions measured in stored serum. Clin Chem 1994;40:1713-1716.
[7] Nanjee NM, Miller NE. Evaluation of long-term frozen storage of plasma for measure- ment of high-density lipoprotein and its subfractions by precipitation. Clin Chem 1990;36:783-788.
K. Evans et al. / Clinica Chimiea Acta 258 (1997) 219 229 229
[8] del Gatto L, Nichols AV, Lindgren FT. Effects of freeze-thawing and storage on ultracentrifugal properties of human serum lipoproteins. Proc Soc Exp Biol Med 1959;101:59-61.
[9] Tiedink HGM, Katan MB. Variability in lipoprotein concentrations in serum after prolonged storage at -20°C. Clin Chim Acta 1989;180:147-156.
[10] Evans K, Mitcheson J, Laker MF. Effect of storage at 4°C and -20°C on lipid, lipoprotein and apolipoprotein concentrations. Clin Chem 1995;41:392 396.
[13] Snipes W, Person S, Keith A, Cupp J. Butylated hydroxytoluene inactivated lipid-con- taining viruses. Science 1975;188:64-66.
[14] Stahler F, Gruber W, Stinshoff K, Roschlav P. Eine praxisgerechte enzymatische Cholesterin-Bestimming. Med Lab 1977;30:29-37.
[15] Takayama M, Itoh S, Nagasaki T, Tanimizu I. A new enzymatic method for the determination of serum choline containing phospholipids. Clin Chim Acta 1977;79:93- 98.
[16] Markwell MAK, Haas SM, Bieber LL, Tolbert NE. A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal Bio~zhem 1978;87:206-210.
[17] Norusis MJ. SPSS for Windows Base System Users Guide. Chicago: SPSS Inc, 1993;396.
[18] Brown SA, Rhodes CE, Dunn K, Gotto AM Jr, Patsch W. Effect of blood collection and processing on radioimmunoassay results for apolipoprotein A-I in plasma. Clin Chem 1988;34:920-924.
[19] Brown SA, Epps DF, Dunn JK, Sharrett AR, Patsch JR, Gotto AM Jr, Patsch W. Effect of blood collection and processing on radioimmunoassay results for apolipo- pro~:ein B in plasma. Clin Chem 1990;36:1662-1666.
[20] Kafbnek S, Bachorik PS, Kwiterovich PO. Measurement of apolipoprotein B in plasma from hyperlipidemic subjects treated with cholestyramine or placebo: differential effects of storage at -70°C. Clin Chem 1989;35:380-383.
