EQUINE VETERINARY JOURNAL
Equine vet. J . (1991) 23 ( 5 ) 353-359
The isolation, characterisation and quantification of the equine plasma lipoproteins T. D. G. WATSON, LYNN BURNS', S. LOVE, C. J. PACKARD' and J. SHEPHERD* Department of Veterinary Medicine, University of Glasgow Veterinary School, Bearsden, Glasgow G6 1 1QH and *Instituteof Biochemistry, Royal Infirmary, Glasgow G4 OSF; UK.
Summary Plasma lipoproteins were isolated from eight Thoroughbred horses and eight Shetland ponies on the basis of particle size by gel filtration chromatography and according to density using rate-zonal ultracentrifugation. Three major classes corresponding to very low density lipoproteins (VLDL), low density lipoproteins (LDL) and high density lipoproteins (HDL) were identified and characterised by their lipid and apolipoprotein compositions. The particle size distributions of each class were determined by electron microscopy and non-denaturing polyacrylamide gradient gel electrophoresis. HDL was found to dominate the equine lipoprotein spectrum, accounting for 61 per cent of the total plasma lipoprotein mass (VLDL 24 per cent, LDL 15 per cent). The VLDL class was isolated as a single population of particles that were triglyceride rich and cholesterol, phospholipid and protein poor. Equine LDL was characteristically cholesterol rich and was found to be polydisperse comprising three subfractions that were discrete with respect to particle size and lipid composition. The HDL class was composed of homogeneous particles that were typically protein rich. Apolipoprotein (apo) B was the major protein of VLDL and LDL, and presented two components on polyacrylamide gel electrophoresis with molecular weights in the region of human apoB-100 and a third in VLDL similar to that of apoB-48. ApoA-I was the predominant protein in equine HDL. Although there were no breed differences in the physical or chemical properties of each lipoprotein class, the Shetland ponies had higher plasma triglyceride and VLDL concentrations than their Thoroughbred counterparts.
(VLDL), low density lipoproteins (LDL) and high density lipoproteins (HDL) have been identified in equine serum (Campbell 1963; Robie, Janson, Smith and O'Connor 1975a). Analytical ultracentrifugation of plasma from two horses subsequently confirmed HDL as the major lipoprotein and suggested that the LDL class was complex (Leat, Northrop, Buttress and Jones 1979). Fractions with the density characteristics of VLDL, LDL and HDL were found in three horses using gradient ultracentrifugation systems with lipid prestaining (Terpstra, Sanchez-Muniz, West and Woodward 1982; Hollanders et a1 1986). Plasma lipoproteins were recently isolated from six Thoroughbred horses using a similar technique and characterised by their chemical composition and particle size distribution on electron microscopy (Le Goff, Nouvelot, Fresnel and Silberzahn 1987; Le Goff et a1 1989). Recognition of a specific hyperlipidaemia in pony breeds (Schotman and Wagenaar 1969; Jeffcott and Field 1985) has stimulated interest in equine lipoprotein metabolism. In the present study, plasma lipoproteins were isolated from 16 healthy horses to establish baseline data for ongoing studies of the pathogenesis of equine hyperlipaemia. Agarose gel filtration chromatography and rate-zonal ultracentrifugation were used to isolate lipoprotein classes, which were characterised by their particle sizes and lipid and apolipoprotein compositions. Plasma concentrations of each class were then quantified. Lipoprotein characteristics of the Shetland pony were compared with those of the Thoroughbred horse to determine whether there are breed differences in basal lipoprotein metabolism that might explain the susceptibility of pony breeds to hyperlipaemia.
Materials and methods
Subjects and samples
PLASMA lipoproteins overcome the difficulties in transporting aqueous insoluble lipids by containing the intensely hydrophobic triglyceride and cholesteryl ester molecules in a core, which is protected from the plasma by a polar coat of free cholesterol, phospholipids and special proteins called apopolipoproteins. Although the horse provided the earliest source of a soluble lipid-protein complex (Macheboeuf 1929) there has been little advancement in our understanding of the equine plasma lipoproteins compared with those of other animal species (Chapman 1980). Alpha, p and pre-P lipoprotein classes with the electrophoretic mobility of very low density lipoproteins
Eight Thoroughbred horses and eight Shetland ponies (two groups of four geldings and four mares) were housed at the University of Glasgow Veterinary School and fed a maintenance diet of hay and proprietary concentrates. Blood samples were collected by jugular venepuncture into EDTA following an overnight fast; plasma was separated by low speed centrifugation and stored at 4°C prior to analysis. Agarose gel filtration chromatography Plasma lipoproteins were isolated from 10 ml plasma after ultracentrifugation at density < I .225 g/ml (39,000 rpm, 1 5 T ,
60 h, Beckman Ti60 rotor) and applied to columns of 6 per cent agarose gel (Bio-Gel A-5m; Bio-Rad Laboratories Ltd.) as described by Rudel, Marzetta and Johnson (1986). The columns were eluted with phosphate buffered saline containing 0.02 per cent sodium azide and the eluant monitored at OD280nm. Fractions comprising the peaks were pooled and concentrated to 2 ml in a micro-concentrator (Amicon Ltd.) and their identity confirmed by compositional analysis. Rate-zonal ultracentrifugation
The two step method of Patsch et a1 (1974) was used first to separate HDL, and then resolve LDL from VLDL, on gradients formed within a Beckman Ti14 rotor (Beckman Industries Ltd.). The density of 50 ml of plasma was adjusted to 1.4 g/ml; it was then applied to a stepped non-linear gradient (density 1.O to 1.4 g/ml) and centrifuged for 18 h at 45,000 rpm at 15°C. The rotor was unloaded at 3,500 rpm and the eluant monitored at OD2gonmand collected into 14 ml fractions. Those fractions (1 to 10) containing the lower density lipoproteins were pooled and concentrated to a volume of 25 ml, adjusted to density 1.3 g/ml and applied to a linear gradient of density 1.O to 1.3 g/ml. The VLDL and LDL classes were separated by centrifugation for 110 mins at 45,000 rpm and 1O'C.
EQUINE VETERINARY JOURNAL
from equine HDL by preparative high performance liquid chromatography (Polacek, Edelstein and Scanu 197 1). The gels were stained with Coomassie Blue and scanned on a video densitometer (Model 620; Bio-Rad Laboratories Ltd.) to calculate the relative percentage of total protein. Quantification of plasma lipoprotein concentrations
Plasma lipid and lipoprotein concentrations were determined by established Lipid Research Clinic protocols (1974). VLDL was isolated from 4 ml plasma by ultracentrifugation at density 1.006 g/ml, LDL was then precipitated from the infranatant by the addition of manganese chloride-heparin, and VLDL and HDL cholesterol concentrations were determined (as above). The LDL cholesterol concentration was calculated by subtraction from the total plasma cholesterol concentration. Statistical analysis
The significance of group differences was determined by the Mann-Whitney test.
Results Agarose gel filtration chromatography
Sequential flotation ultracentrifugation
VLDL, intermediate density lipoproteins (IDL), LDL and HDL were sequentially isolated from 4 ml plasma at the density limits of 1.006, 1.019, 1.063 and 1.21 g/ml (Havel, Eder and Bragdon 1955) using a Beckman Ti50.3 rotor (39,000 rpm, 4'C) to evaluate breed differences in chemical composition and determine particle size by electron microscopy. Determination of particle size distributions
The equine plasma lipoproteins consistently appeared as three discrete peaks with elution volumes that corresponded to human VLDL, LDL and HDL (Fig 1). A fourth, partially resolved peak corresponding to IDL was present in varying degrees in the tail of the VLDL elution profile. HDL was the major lipoprotein class in all animals. Compositional analysis of the fractions comprising the three major peaks confirmed their identity (Table 1). The VLDL class was characteristically triglyceride rich (54 per cent) and poor in cholesterol (16 per cent), phospholipid (14 per cent) and protein (16 per cent)
Lipoprotein fractions were dialysed against a 0.125 M ammonium acetate, 2.6 mM ammonium carbonate buffer, pH7.4 and negatively stained with phosphotungsic acid prior to electron microscopy (Forte and Nordhausen 1986). The particle size distribution of equine HDL and LDL isolated by rate-zonal ultracentrifugation was determined by non-denaturing electrophoresis on prepared gradient gels of 2 to 16 per cent (for LDL) and 4 to 30 per cent (for HDL) polyacrylamide (PAA 2/16, 4/30; Pharmacia AB) as described by Griffin et a1 (1990). The gels were stained with Coomassie Blue; a high molecular weight marker (HMW; Pharmacia AB) and a 10 per cent solution of latex beads (diameter 38 nm; Dow Chemical Ltd.) were included to calibrate particle size.
Fig I : Elution profile of plasma lipoproteins (d > 1.225 glml) from pony a Shetland separated by gel
Determination of lipid and apolipoprotein compositions
Triglyceride and total cholesterol were measured using reagent kits (Boehringer Mannheim GmbH) on an automated discrete analyser (BMkIitachi 717, Boehringer Mannheim GmbH). Free cholesterol and phospholipid were determined using enzymatic colorimetric kits (Boehringer Mannheim GmbH). Protein was assayed by the modified method of Lowry et a1 (Peterson 1977). Percentage apoB in VLDL and LDL was estimated by the tetramethylurea (TMU) precipitation technique described by Kane, Sata, Hamilton and Havel (1975). The apolipoproteins in each class were separated by electrophoresis on 3.5 per cent SDS-glycerol polyacrylamide gels (SDS-PAGE) according to the method of Maguire, Lee and Connelly (1989). Low and high molecular weight markers (LMW and HMW-SDS; Pharmacia AB) were included in each run. ApoA-I, apoA-I1 and the C peptides were partially purified
Elution volume (ml)
filtration chromatography on a column of 6 per cent agarose and monitored at OD280nm
TABLE 1: Compositional analysis of the major equine lipoprotein classes VLDL, LDL and HDL isolated by agarose gel filtration chromatography
Mean ksd of 16 animals. CE: cholesteryl esters; TG: triglyceride; FC: free cholesterol; PL: phosholipid; ND: not detectable
EQUINE VETERINARY JOURNAL
TABLE 2: Agarose gel filtration chromatography; analysis of lipoprotein absorbency profile parameters for eight Thoroughbred horses (TB) and eight Shetland ponies (SP) VLDL
Peak height (OD),
TB SP Elution volume (ml) TB SP Fraction width (ml) TB SP
Lipoprotein class LDL
0.17f 0.11 0.13f 0.05 0.43f 0.1 9’ 0.1 5 f 0.04 80.4f 2.1 119.4f 2.5 79.1 f1.4 117.1 f1.5 10.4f 0.8 28.4 f 2.8 10.9f 0.8 31.8f 2.1
1.80 f 0.25 1.88f 0.31 157.6 f 2.6 155.6f3.4 34.7f 3 36.1 f 3
Mean k sd. ‘P < 0.01
Elution volume (ml) Fig 2: Representative elution profile of equine HDL isolated by ratezonal ultracentrifugation in a stepped gradient non-linear NaBr gradient (density 1 .O to 1.4 glml) and unloaded through an optical meter reading OD280nm
Elution volume (ml) Fig 3: Separation of equine LDL by rate-zonal ultracentrifugation in o linear NaBr gradient (density 1 .O to 1.3 glml), demonstrating three LDL overlapping subractions along the density gradient; LDLl , LDL2 and LDL3
relative to LDL and HDL. There was progressive triglyceride depletion and enrichment in cholesterol as particle size decreased into the LDL class, which contained 6 per cent triglyceride and 47 per cent total cholesterol. The HDL class also was triglyceride poor (9 per cent), but protein rich (43 per cent) with an intermediate cholesterol content (22 per cent). Analysis of the profile parameters (Table 2) for each animal revealed no breed differences in the particle size (elution
Fig 4: Electron micrographs (200,000X) of intact VLDL [top]. LDL [centre] and HDL [bottom] particles prepared by flotation ultracentrifugation and negatively stained with phosphotungsic acid; diameter VLDL, 30 to 60 nm; LDL 20 to 30 nm and HDL, 5 to 8 nm
volume) or dispersity (fraction width) of each lipoprotein class. Significantly, the apparent mass of VLDL (peak amplitude) was greater in the Shetland pony than the Thoroughbred (P