Atherosclerosis 157 (2001) 211 220 www.elsevier.

com/locate/atherosclerosis

The atherogenic lipoprotein phenotype: small dense LDL and lipoprotein remnants in nephrotic range proteinuria
Christopher J. Deighan a,b,*, Muriel J Caslake b, Michael McConnell b, J. Michael Boulton-Jones a, Christopher J. Packard b
b a Renal Unit, Walton Building, Glasgow Royal Inrmary, Castle St., Glasgow, G4 0SF, UK Department of Pathological Biochemistry, Glasgow Royal Inrmary, Castle St., Glasgow, G4 0SF, UK

Received 14 July 1999; received in revised form 14 July 2000; accepted 19 October 2000

Abstract The dyslipidaemia in nephrotic-range proteinuria is believed to contribute to the increased atherogenesis associated with the condition. Excess small dense low density lipoprotein (LDLIII) contributes to this risk. Lipoprotein remnants (RLP) may also be implicated but have not been studied in this population. We measured the plasma concentration of low density lipoprotein (LDL) subfractions (by density gradient ultracentrifugation), RLP (by immunoafnity gel), very low density lipoprotein (VLDL) subfractions, post heparin lipases and cholesteryl ester transfer protein (CETP) activity in 27 patients with glomerular disease and albuminuria \2.0g. These were compared with 27 age and sex matched controls. Proteinuric patients had increased LDLIII concentration (patients 182 (84:267) vs. controls 31 (27:62); PB 0.0001) with reduced lighter LDLI (36 (24:43) vs 69 (46:101); PB0.0005) and LDLII (124 (79:220) vs 178 (129:236); P B 0.04, all mg/dl, median + interquartile range). RLP-cholesterol (RLP-C) and triglyceride (RLP-TG) were increased in proteinuric patients (RLP-C 18.9 (11.0:26.9) vs 7.7 (6.0:8.8); P B 0.0001, RLP-TG 35.8 (11.8:54.7) vs. 7.2 (4.3:10.0); PB 0.0001, all mg/dl). Increased LDLIII and RLP were independent of renal function. VLDL1 and VLDL2 concentrations were increased by 258 and 260% (both PB 0.0001). CETP activity was increased by 46% (PB0.005). Lipoprotein and hepatic lipase activities did not differ from control values. LDLIII concentration (r 2 = 45.7%, PB 0.001), RLP-C (r 2 =85.2%, P B0.001) and RLP-TG (r 2 = 87.5%, P B0.001) all correlated positively with plasma triglyceride. Moreover, increased LDLIII was associated with both RLP-C (r 2 = 31.3%, PB 0.002) and RLP-TG (r 2 = 33.6%, PB0.002). Excess LDLIII and RLP are present in nephrotic-range proteinuria and add to the spectrum of cardiovascular risk factors present in proteinuric patients. Increases in LDLIII and RLP are closely related to plasma triglyceride. The association between excess RLP and LDLIII suggests that RLP contribute to the increased atherogenicity attributed to the atherogenic lipoprotein phenotype. 2001 Elsevier Science Ireland Ltd. All rights reserved.
Keywords: Low-density lipoprotein; Small dense LDL; Remnant lipoproteins; Atherogenic lipoprotein phenotype; Nephrotic range proteinuria; Cardiovascular risk

1. Introduction Nephrotic range proteinuria is associated with abnormal lipoprotein metabolism and increased cardiovascular risk [1]. Quantitative and qualitative changes in lipids and lipoproteins exist. Total cholesterol and low density lipoprotein (LDL) cholesterol (LDL-C) are invariably raised [2,3]. Triglycerides may also be increased [4] and very low density lipoprotein cholesterol
* Corresponding author. Tel.: + 44-141-2115049; fax: +44-1412114843. E-mail address: cdeighan@compuserve.com (C.J. Deighan).

(VLDL-C) is raised [2,3], even in patients with proteinuria, normal serum albumin and normal renal function [5]. This dyslipidaemia results from a combination of increased production and reduced clearance of lipoproteins. VLDL is overproduced, lipolysis of VLDL to intermediate density lipoprotein (IDL) and LDL is delayed, and receptor mediated clearance of LDL reduced [2]. There is no clear pattern of the effect of proteinuria on activity of lipoprotein and hepatic lipase [2,3]. Lipoproteins are generally classied according to their density, however within each density range they

0021-9150/01/$ - see front matter 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 2 1 - 9 1 5 0 ( 0 0 ) 0 0 7 1 0 - 3

212

C.J. Deighan et al. / Atherosclerosis 157 (2001) 211220

exhibit heterogeneity [6]. LDL can be subdivided into large, light LDLI (d 1.025 1.034 g/ml) and LDLII (d 1.0341.044 g/ml), and small dense LDLIII (d 1.044 1.063 g/ml). Similarly VLDL can be divided into large, light VLDL1 (Sf 60 400) and smaller, denser VLDL2 (Sf 2060) [6]. Increasing numbers of studies have demonstrated that excess small dense LDL is associated with increased risk of coronary artery disease [7]. A LDLIII concentration \100 mg/dl confers a sevenfold increase in the risk of myocardial infarction [8] and patients with coronary artery disease have smaller LDL size [9]. Normal populations require raised triglycerides (in the form of triglyceride rich VLDL1) and adequate activity of hepatic lipase and cholesteryl ester transfer protein (CETP) to form atherogenic levels of LDLIII [6]. Thus the combination of hypertriglyceridaemia, increased small dense LDL and also low HDL are physiologically linked and known as the atherogenic lipoprotein phenotype (ALP) [6,7]. We have recently shown that patients with nephrotic range proteinuria have a concentration of LDLIII associated with increased risk of coronary heart disease (CHD) [10]. Recent evidence suggests that excess lipoprotein remnants (RLP) are atherogenic [11] and a risk factor for coronary artery disease [12]. Excess RLP are a characteristic nding in patients with hypertriglyceridaemia [13] and therefore may contribute to the increased cardiovascular risk attributed to the atherogenic lipoprotein phenotype. RLP have not previously been studied in patients with nephrotic range proteinuria however given the delayed clearance of triglyceride-rich lipoproteins found in this population, excess RLP are likely to be present and may contribute to the increased CHD risk seen in nephrotic dyslipidaemia. The aim of this study was to investigate the appearance of the atherogenic lipoprotein phenotype in patients with nephrotic range proteinuria, examining factors determining formation of both small dense LDL and remnant lipoproteins (RLP). At the same time, this would allow us to conrm our previous nding of excess LDLIII in this population, clarify the inuence of renal dysfunction, plasma and urinary albumin on LDLIII and RLP, and examine any relationship between lipoprotein remnants and small, dense LDL in proteinuric renal disease.

total protein loss of \ 3.5g/24 h); (b) serum creatinine B 300 umol/l; and (c) glomerular disease. They were compared to 27 age and sex matched controls. Patients suffering from other diseases or on treatment that might inuence their lipid prole were excluded, specically patients with underlying liver disease, diabetes mellitus, amyloid, any neoplastic disorder, systemic lupus erythematosis or taking thiazide diuretics, fat soluble beta blockers, corticosteroids or other immunosuppressive agents. Treatment with other antihypertensives or diuretics was permitted. Patients receiving lipid lowering therapy had their lipid lowering medication stopped for a period of 4 weeks prior to inclusion in the study. All patients had a biopsy proven diagnosis. They were compared to 27 age and sex matched controls who were either relatives of patients attending the renal unit or laboratory staff members. All controls were free of acute illness with normal serum creatinine, albumin and liver function tests. Any patient or control with a bleeding diathesis or with a recent history of peptic ulceration was excluded. All patients and controls gave written consent before participating. The study was approved by the Ethics Committee of Glasgow Royal Inrmary. Patients and controls attended after an overnight fast. An intravenous cannula was inserted and samples were taken for lipids and lipoproteins, LDL and VLDL subfractions, serum creatinine and albumin. Heparin was then given (70 units/kg body weight, intravenous) followed by further sampling 10 min later for hepatic and lipoprotein lipase measurement. All patients collected a 24 h urine for analysis of urinary albumin excretion.

2.2. Methods
Plasma cholesterol, triglyceride, VLDL cholesterol, LDL cholesterol and HDL cholesterol were analysed by a modication of the standard Lipid Researchs Clinics (LRC) protocol. VLDL1 (Sf 60400), VLDL2 (Sf 2060) and IDL (Sf 1020) were isolated from plasma by a modication of the cumulative-gradient ultracentrifugation procedure [14]. LDL was isolated from fresh plasma by sequential ultracentrifugation. The triglyceride, free cholesterol, cholesteryl ester, phospholipid and protein content of the lipoproteins was assayed and lipoprotein concentrations were calculated as the sum of these components. Three LDL subfractions were isolated from fresh plasma by nonequilibrium density gradient ultracentrifugation as previously described [15]. Lipoprotein and hepatic lipase activities were measured in post-heparin plasma incubated with a gum arabic-stabilized triglyceride emulsion containing glycerol tri(1-14C) oleate at a specic activity of 30 mCi/mmol triglyceride fatty acids as substrate [16]. For the specic assay of lipoprotein lipase (LPL), plasma was preincubated with sodium

2. Subjects and methods

2.1. Subjects
Twenty-seven patients (22 males, ve females) were recruited consecutively from the out-patient clinic of the Glasgow Royal Inrmary Renal Unit, according to the following inclusion criteria (a) urinary albumin excretion \ 2.0g/24 h (approximately equivalent to a

C.J. Deighan et al. / Atherosclerosis 157 (2001) 211220

213

dodecyl sulfate (SDS) to inhibit hepatic lipase (HL), and pooled pre-heparin plasma was added to the incubation as a source of cofactor apoCII to activate the LPL enzyme. HL was assayed in the presence of 1.0 mol/l NaCl to inactivate LPL. Enzyme activities are expressed in mmol of fatty acids released per hour per ml of plasma. CETP was assayed using HDL with uorescein labelled cholesteryl ester using the plasma from patients and controls as the VLDL acceptor (Roar Biomedical, New York, USA). The activity is expressed as pmol cholesteryl ester transferred per ml of plasma. The cholesterol and triglyceride in remnant like particles (RLP-C, RLP-TG) were measured using a diagnostic reagent kit from Japan Immunoresearch Laboratories (JIMRO), Takasaki, Japan. Remnant lipoproteins were separated from other lipoproteins by an immunoafnity gel containing monoclonal antibodies to human apoB100 and human apoA1 [17]. RLP measurements were calculated on frozen samples (freezing may result in elevation of RLP-C, but elevation is minimal if the sample is frozen and thawed quickly, and well mixed with the monoclonal antibodies for 30 min before analysis stability data from manufacturers). The cholesterol and triglyceride contents of the unbound fraction were measured using sensitive enzymatic spectrophotometric assays. The coefcient of variation of the assays was B15%.

creatinine in the patient group was 120 mmol/l (IQR 100:180), with the creatinine clearance (CrCl) greater than 50 ml/min in 20 patients (calculated using the Cockcroft and Gault formula [18]), and greater than 70 ml/min in 15 patients. In the patient group median urinary albumin excretion was 2.8 g/24 h (32:40), approximately equivalent to 4.7g of proteinuria per 24 h. Serum albumin was well preserved median 37g/l (IQR 32:40). Six patients had a serum albumin B 32 g/l. The diagnoses were membranous nephropathy (eight patients) IgA nephropathy (seven), focal and segmental glomerulosclerosis (four), mesangiocapillary glomerulonephritis (three) and minimal change (three). In one patient the glomerular disease was secondary to Alports syndrome. One patient had glomerular disease secondary to nailpatella syndrome.

3.2. Lipids, lipoproteins and post-heparin lipase acti6ity

The results obtained for lipids, lipoproteins and postheparin lipase activity are shown in Table 1. Plasma cholesterol, triglyceride and VLDL-C, were all higher in patients with proteinuria compared with controls (all PB 0.0001), LDL-C was similarly raised with HDL-C reduced in the patients (both PB 0.003). No difference in activity of either lipoprotein or hepatic lipase was observed. There was a three to fourfold increase in total VLDL in the circulation (PB0.001), due to an increase in the concentration of both VLDL1 and VLDL2 subfractions in proteinuric patients (both PB 0.001). IDL concentration in the patients was twice that of controls (PB 0.0001). Total plasma LDL concentration was also increased in proteinuric patients (PB 0.02). Twenty-four hour urinary albumin loss correlated positively with plasma cholesterol (r 2 = 22%, PB 0.02), LDL-C as measured by the LRC protocol (r 2 = 15%, PB 0.05) and IDL concentration (r 2 = 17%, PB0.04). No correlation with plasma LDL concentration was observed. An inverse correlation was seen between 24 h urinary albumin loss and HDL-C (r 2 = 21%, PB0.02). Low serum albumin was associated with increasing plasma cholesterol (r 2 = 41%, P B0.001), LDL-C (r 2 = 41%, PB 0.001), IDL concentration (r 2 = 27%, PB 0.006) and total LDL concentration (r 2 = 19%, PB 0.03). There was no observed relationship between either plasma creatinine or CrCl and any lipid parameter in the population studied.

2.3. Statistics
Statistical analyses were performed using MINITAB version 13 for Windows (Minitab Inc.). Factors that were not normally distributed were subjected to log transformation. These included: BMI, plasma TG, VLDL-C, HDL-C, LPL, HL, total VLDL and VLDL subfractions, total LDL and LDL subfractions and lipoprotein remnants. Results are shown as medians and interquartile ranges (IQR). Samples were compared using the twosample t-test. Patient subgroups were compared to the control group using analysis of variance and Fishers individual error rate when the overall F statistics were signicant. Linear regression analysis was performed to identify signicant correlations. Multivariate analysis was performed using a general linear model with a P B 0.05 used to determine statistical signicance. Results are expressed as a regression coefcient with 95% condence intervals and an r 2 value, where r 2 determines the independent contribution of the variable to the variation in the parameter analysed.

3.3. LDL subfractions, CETP acti6ity and remnant lipoproteins

3. Results A marked shift in LDL subfraction prole was observed in the patients with proteinuria. Large light LDLI and LDLII were both reduced, with a marked increase in plasma LDLIII concentration (PB 0.0001, Table 2). LDLIII concentration was greater than

3.1. Anthropometry, diagnoses and renal function

The age and sex matched patient and control groups had similar body mass index (BMI; Table 1). The median

214

C.J. Deighan et al. / Atherosclerosis 157 (2001) 211220

Table 1 Anthropometry, lipids, lipoproteins and lipase activity (Median+IQR) Controls Age (years) Sex (M:F) Body Mass Index (kg/m2) Creatinine (mmol/l) Cholesterol (mmol/l) Triglyceride (mmol/l) VLDL-C (mmol/l) LDL-C (mmol/l) HDL-C (mmol/l) Total VLDL concentrationa (mg/dl) VLDL1 concentrationa (mg/dl) VLDL2 concentrationa (mg/dl) IDL concentrationa (mg/dl) Total LDL concentrationa (mg/dl) Lipoprotein lipase (mmolFA/ml/h) Hepatic lipase (mmolFA/ml/h)
a b

Patients 48 (29:62) 22:5 26.5 (25.1:30.2) 120 (100:180) 6.8 (5.7:7.5) 2.3 (1.6:3.3) 1.0 (0.5:1.4) 4.3 (3.7:5.5) 1.0 (0.8:1.3) 236 (172:364) 156 (57:212) 109 (75:140) 80 (63:113) 351 (298:441) 3.7 (2.9:5.6) 16.2 (10.1:24.3)

Pb nsd nsd nsd B0.003 B0.001 B0.001 B0.001 B0.003 B0.003 B0.0001 B0.0001 B0.0001 B0.0001 B0.02 nsd nsd

45 (30:60) 22:5 24.5 (22.8:28.6) 100 (95:120) 5.2 (4.6:5.8) 1.0 (0.9:1.3) 0.3 (0.2:0.5) 3.5 (2.8:4.0) 1.3 (1.1:1.7) 79 (52:113) 43 (24:74) 30 (21:43) 44 (34:60) 292 (252:349) 4.0 (3.4:5.4) 13.8 (9.6:16.1)

Lipoprotein concentration = the sum of triglyceride, free cholesterol, cholesteryl ester, phospholipid and protein content of the lipoproteins. Statistical comparison by Students t-test.

100mg/dl in 74% of the patients studied, compared with 4% of controls. Similarly, the percentage of plasma LDL that was in the form of LDLIII (%LDLIII) was increased in the patients (P B 0.0001), with a corresponding decrease in the %LDLI (P B 0.0001) and %LDLII (P B 0.0003, Table 2), suggesting that the difference in LDL prole was independent of the total plasma LDL concentration. CETP activity was increased in the proteinuric patients (P B 0.005, Table 2). Both plasma VLDL1 and VLDL2 subfractions were found to be cholesterol enriched and triglyceride deplete compared with controls, whereas LDL was cholesterol deplete and triglyceride enriched in proteinurics (Table 3). There was a two to threefold increase in RLP-Cholesterol and a vefold increase in RLP-Triglyceride (both P B0.0001, Table 2) in the proteinuric compared with control groups. No signicant correlation was observed between 24 h urinary albumin loss and plasma LDLIII concentration, CETP activity or remnant lipoproteins.

3.4. Effect of renal function and serum albumin

Marked heterogeneity in terms of renal function and serum albumin was observed amongst the proteinuric patients. As a result, subgroup analysis was performed on the patients with a creatinine clearance (CrCl) \50 ml/min (n =19), and on those with a serum albumin \ 32 g/l (n =21). The age and sex distribution of patients with a CrCl \50 (median 45 years (IQR 45:62), 15 males: four females) was similar to that of controls. Median urinary albumin was

2.7 g/24 h (IQR 2.1:4.9), plasma LDLIII concentration, %LDLIII, CETP activity, RLP-C and RLP-TG were all markedly elevated compared with the control group (LDLIII 195 mg/dl (73:288) P B 0.0001, %LDLIII 57.2 (17.7:74.6) PB 0.0001, CETP 63.2 pmolCE/ml (49.9:66.9) P B 0.008, RLP-C 18.9 mg/dl (9.6:26.4) PB 0.0002, RLP-TG 35.8 mg/dl (13.3:47.0) PB 0.0001 (all median+ IQR, P vs. controls). No correlation was observed between renal function (using either CrCl or reciprocal of creatinine) and CETP activity, remnant lipoproteins or plasma LDLIII (cont. centration or percentage). A similar subgroup analysis was performed using the 21 patients with a serum albumin \ 32 g/l. Again, age and sex distribution was similar to controls (median 48 years (IQR 31:62), 17 males: four females), with the proteinuric but normoalbuminaemic patients having signicant increases in plasma LDLIII concentration (178 mg/dl (69:274) P B 0.0001), %LDLIII (57.4 (21:72) P B 0.0001), RLP-C (18.8 mg/ dl (10.3:25.8) P B 0.0001) and RLP-TG (35.8 mg/dl (11.6:54.2) PB 0.0001), but not CETP (55.2 pmolCE/ ml (40.9:66.4) P B 0.08), (all median+ IQR, P vs controls). No relationship was observed between serum albumin and LDLIII, CETP or remnant lipoproteins. Amongst the 15 patients with a CrCl \50 and a normal serum albumin, the median LDLIII concentration was 187 mg/dl (49:288), 57.4% (17.7:74.6) of the LDL present was in the form of LDLIII, with a RLP-C of 18.9 mg/dl (9.6:25.4) and a RLP-TG of 35.8 mg/dl (9.5:46.0); all median+ IQR.

Table 2 LDL subfractions, remnant lipoproteins and CETP (Median+IQR) Controls (n =27) LDLI conc.a (mg/dl) LDLII conc a (mg/dl) LDLIII conc. a (mg/dl) %LDLI %LDLII %LDLIII CETP (pmolCE/ml) RLP-C (mg/dl) RLP-TG (mg/dl)
a

C.J. Deighan et al. / Atherosclerosis 157 (2001) 211220

Patients all (n =27) 36 (24:43) 124 (79:220) 182 (84:267) 9.4 (6.9:12.4) 35.6 (20.8:63.4) 57.1 (22.0:72.3) 62.7 (48.9:69.0) 18.9 (11.0:26.9) 35.8 (11.8:54.7)

Pb (vs. controls) B0.0005 B0.04 B0.0001 B0.0001 B0.0003 B0.0001 B0.005 B0.0001 B0.0001

Patients TG 52.3 mmol/l (n = 14) 36 (26: 64) 198 (122:272) 100 (32:176) 12.0 (10.0:19.7) 61.2 (37.9:68.6) 23.1 (12.0:54.6) 53 (27:63) 11.5 (7.2:13.8) 12.5 (9.1:16.6)

Pc (vs controls) Nsd Nsd B0.005 B0.001 Nsd B0.005 Nsd B0.02 B0.007

Patients TG \2.3 mmol/l (n= 13) 31 (11: 39) 82 (47:126) 264 (218:299) 8.4 (4.3:9.1) 20.8 (16.9:30.8) 72.3 (60.5:77.7) 67 (55:90) 26.9 (24.6:49.4) 54.7 (43.7: 81.9)

Pc (vs controls) B0.001 B0.001 B0.001 B0.001 B0.001 B0.001 B0.001 B0.001 B0.001

69 (46:101) 178 (129:236) 31 (27:62) 19.4 (15.4:33.5) 62.9 (53.3:69.5) 10.6 (8.5:18.6) 43.0 (37.2:54.2) 7.7 (6.0:8.8) 7.2 (4.3:10.0)

Lipoprotein concentration = the sum of triglyceride, free cholesterol, cholesteryl ester, phospholipid and protein content of the lipoproteins. Patients (all) vs controls: comparison by Students t-test. c Patients TG 52.3 mmol/l and \2.3 mmol/l vs controls: comparison by ANOVA and Fishers individual error rate.
b

215

216 Table 3 Lipoprotein compositions (Median+IQR) Controls VLDL1% Cholesterola VLDL1% Triglyceridea VLDL2% cholesterola VLDL2% triglyceridea IDL% cholesterola IDL% triglyceridea LDL% cholesterola LDL% triglyceridea 7.0 (5.7:8.6) 67.4 (65.1:70.4) 18.8 (16.6:21.5) 44.1 (38.6:46.5) 36.0 (33.7:38.5) 14.3 (12.0:17.8) 39.8 (38.5:40.4) 6.5 (6.0:7.1)

C.J. Deighan et al. / Atherosclerosis 157 (2001) 211220

Patients 10.6 (7.9:12.2) 64.8 (62.0:67.1) 24.7 (21.7:27.5) 37.5 (33.4:42.0) 36.1 (34.0:38.8) 16.2 (13.4:19.6) 38.4 (36.5:39.9) 8.2 (6.7:9.1)

Pb B0.0001 B0.0001 B0.0001 B0.03 nsd nsd B0.02 B0.004

a % Cholesterol or Triglyceride =% of lipoprotein concentration that is cholesterol or triglyceride. b Statistical Comparison by Students t-test.

3.5. Factors determining LDLIII and lipoprotein remnants in proteinuric patients

Plasma LDLIII concentration correlated with plasma triglycerides, CETP activity (both Fig. 1), and to concentrations of VLDL1 (r 2 =29.7%, P B 0.005), VLDL2 (r 2 =39.1%, PB 0.001) and total LDL (r 2 = 24.3%, PB 0.01). An inverse relationship was seen with HDL-C (r 2 =38.0%, P B0.005). No relationship was observed between LDLIII and either lipoprotein lipase or hepatic lipase activity (Fig. 1). Multivariate analysis revealed that plasma triglycerides and total LDL concentration were independent predictors of LDLIII concentration (Table 4). %LDLIII was related to plasma triglyceride (r 2 =54.9%, P B 0.001) and VLDL1 (r 2 =57.9%, P B0.001) with a weaker relationship with VLDL2 (r 2 =20.2%, P B 0.05) and CETP activity (r 2 = 20.2%, P B 0.05). An inverse relationship was seen with HDL-C (r 2 =32.9%, P B 0.05) with no correlation seen with total LDL or lipase activity. On multivariate analysis plasma triglyceride was the only independent predictor of %LDLIII (Table 4). RLP-C and RLP-TG were closely correlated with plasma triglyceride (RLP-C r 2 =85.2% (Fig. 2), RLPTG 87.5%, both P B0.001), VLDL1 (RLP-C r 2 = 68.8% (Fig. 2), RLP-TG 85.4%, both P B0.001), VLDL2 (RLP-C r 2 =27.6% P B 0.005, RLP-TG 18.3% PB 0.05), and plasma CETP activity (RLP-C r 2 = 46.0%, RLP-TG 39.3%, both P B0.001, Fig. 2). An inverse correlation was noted with HDL-C (RLPC r 2 = 39.2%, RLP-TG r 2 =45.0%, both P B 0.001). No relationship was observed between remnant

Fig. 1. Plasma triglyceride, CETP activity and hepatic lipase activity vs. LDLIII concentration in proteinuric patients.

lipoproteins and either lipase activity or total LDL concentration. On multivariate analysis plasma triglyceride was the only predictor of RLP-C and RLP-TG concentration (Table 4). A clear relationship was also observed between LDLIII and RLP. Plasma LDLIII concentration was associated with both RLP-C (r 2 = 31.3% PB0.003, Fig. 3) and RLP-TG (r 2 = 33.6% PB0.003), with a closer relationship observed between %LDLIII and both RLP-C (r 2 = 46.0% PB 0.001, Fig. 3) and RLPTG (r 2 = 55.5% PB 0.001).

C.J. Deighan et al. / Atherosclerosis 157 (2001) 211220

217

3.6. Plasma triglyceride, LDLIII and remnant lipoproteins

NCEP guidelines currently dene a raised triglyceride as being \200 mg/dl (2.3 mmol/l) [19]. Given the close relationship observed between plasma triglyceride and both LDLIII and RLP, subgroup analysis was performed comparing patients with plasma triglyceride 5 or \ 2.3 mmol/l with controls (Table 2). The patient subgroup with plasma triglyceride \2.3 mmol/l was found to have increased %LDLIII, with increased concentration of LDLIII, RLP-C and RLP-TG, and excess CETP activity. However compared to controls the patients with plasma triglycerides B2.3 mmol/l were also found to have excess %LDLIII, and increased plasma concentration of LDLIII, RLP-C and RLP-TG (Table 2). Median CETP activity, although increased, varied widely and therefore was not statistically different to controls.

4. Discussion This study demonstrates that remnant lipoproteins are present in excess in patients with nephrotic range proteinuria. The data also conrms the original ndings of our pilot study suggesting that small dense LDL is present in quantities that are associated with signicantly increased cardiovascular risk in normal populations. This analysis has been extended, demonstrating that the excess of both RLP and LDLIII is not related to mild renal impairment and that no direct relationship was discernible with either urinary or serum albumin. In addition plasma triglyceride is demonstrated to be the most important factor determining plasma concentration of LDLIII and RLP, with increased plasma concentration of these atherogenic particles present in patients who would be regarded as having normal plasma triglycerides [19]. Comparing the RLP results to those obtained from the Framingham Heart Study, demonstrates that the proteinuric patients exhibit RLP-C and RLP-TG concentrations, twothree-fold higher than the reference value for males [20]. Note that the control RLP-C data corresponded closely to reference values [20]. It is

recognised that younger patients and pre-menopausal women have lower RLP levels. Therefore the difference between the control RLP-TG data and reference male data [20] is likely to result from the younger mean age of the patients (compared to the Framingham Offspring Study) and the presence of four premenopausal women. In keeping with other dyslipidaemic populations, plasma triglyceride was the most important determinant of RLP-C and RLP-TG concentration [13]. It is recognised that this excess plasma triglyceride results from increased plasma concentration of both VLDL1 and VLDL2, which is likely to be due to impaired clearance of these triglyceride-rich lipoproteins [2,21,22]. It is, therefore, likely, that the impaired clearance of VLDL1 and VLDL2 is also related to the excess of RLP observed. In-vitro lipase activity was not reduced in the proteinuric group, therefore it was perhaps not surprising that there was no association between lipase activity and RLP concentration. Rather, in this condition, it has been suggested that the defective clearance of VLDL may result from deciencies in plasma or VLDL apolipoproteins [3]. The current theory describing the formation of atherogenic levels of LDLIII requires the presence of excess triglycerides and adequate levels of hepatic lipase (found in males and post-menopausal females). Triglyceride is passed via CETP from VLDL1 to LDL in exchange for cholesteryl ester. This results in the formation of a triglyceride enriched LDL particle which is hydrolysed by hepatic lipase, shrinking the particle and forming small dense LDL [6]. The results obtained in this study suggest that LDLIII formation in proteinuric patients follows similar lines. Increased plasma triglyceride is related to both plasma LDLIII concentration and %LDLIII. Plasma CETP activity is increased and correlated with the concentration and percentage of LDLIII, whilst VLDL1 concentration is increased and is cholesteryl ester enriched. In addition, the LDL present is triglyceride rich, a characteristic nding for small dense LDL [6]. It is likely that hepatic lipase activity was not rate limiting due to the extent of the hypertriglyceridaema. Thus both plasma triglyceride and CETP activity contribute to the formation of LDLIII in proteinurics, with the multivariate analysis suggesting that plasma triglyceride is by far the most important predictor.

Table 4 Results of multivariate analysis to assess factors determining LDLIII and remnant lipoproteins Response variable LDLIII concentration % LDLIII RLP-Cholesterol RLP-Triglyceride
a

Determinant variable Plasma Plasma Plasma Plasma Plasma triglyceride LDL concentration triglyceride triglyceride triglyceride

Regression coefcient+95% condence intervalsa 22.58 0.35 7.57 9.54 37.93 (4.14:41.02) (0.08:0.62) (3.31:11.83) (8.39:10.69) (33.84:42.02)

r 2% 17.2 20.3 54.9 92.0 93.0

Pa B0.02 B0.02 B0.001 B0.001 B0.001

Statistical comparison by General Linear Model.

218

C.J. Deighan et al. / Atherosclerosis 157 (2001) 211220

[26] and can promote lipid accumulation by macrophages [11]. Prospective and retrospective studies have established that increased small dense LDL is associated with enhanced cardiovascular risk [8,9] and patients with an LDLIII mass of \ 100 mg/dl have a seven-fold increase in the risk of myocardial infarction [8]. Using this criteria, 20 of the 27 patients studied have a high risk of developing a myocardial infarction, irrespective of other cardiovascular risk factors. The atherogenicity of small dense LDL is thought to be related to decreased LDL receptor afnity resulting in longer residence time in plasma, increased uptake by the scavenger receptor, increased susceptibility to oxidation, increased ltration by the endothelium and greater afnity for binding to intimal proteoglycans [6]. NCEP guidelines dene a normal plasma triglyceride as B 200 mg/dl (2.3 mmol/l) [19]. It was therefore interesting to note that despite the close correlation between plasma triglyceride and both LDLIII and RLP, the subgroup of patients with normal plasma triglycerides continued to have increased levels of both

Fig. 2. Plasma triglyceride, VLDL1 concentration and CETP activity vs. RLP-C in proteinuric patients.

There is increasing evidence to link both lipoprotein remnants and small dense LDL with coronary artery disease. RLP-C levels are increased in men with coronary artery disease [12]. High RLP-C is a predictor for coronary artery stenosis [23] and sudden cardiac death [24]. RLP induces platelet aggregation [25], is associated with coronary artery endothelial cell dysfunction

Fig. 3. LDLIII concentration and %LDLIII vs. RLP-C: Patients.

C.J. Deighan et al. / Atherosclerosis 157 (2001) 211220

219

LDLIII and RLP. In males it is recognised that a plasma triglyceride \130 mg/dl (1.5 mmol/l) is the threshold above which atherogenic levels of LDLIII are formed. Moreover, excess remnant lipoproteins have been demonstrated in men with coronary artery disease and plasma triglyceride B200 mg/dl [27]. The data presented here supports the theory that excess atherogenic lipoprotein particles are produced at levels of plasma triglyceride previously regarded as normal. Patients with end-stage renal failure have increased small dense LDL [28] and RLP [29]. However we found no evidence in this study to suggest that mild renal dysfunction contributed to the increase in LDLIII or RLP observed in the proteinuric patients. The lack of correlation between either LDLIII or RLP and both plasma or urinary albumin in this population is also noteworthy. In conclusion, the central role of plasma triglycerides in determining the extent of RLP and LDLIII formation in patients with proteinuria is clearly demonstrated in this study. To our knowledge LDLIII and RLP have not been analysed in the same study population, however the close correlation between plasma triglycerides, RLP and LDLIII, suggests that RLP may contribute to the increased cardiovascular risk attributed to the atherogenic lipoprotein phenotype [7]. We suggest that the denition of the syndrome known as the atherogenic lipoprotein phenotype may require modication to include RLP (i.e. in addition to small dense LDL, mild hypertriglyceridaemia and low HDL). The nding of excess LDLIII and RLP in the patients with proteinuria, highlights the need to clarify the nature of the defective clearance of triglyceride rich lipoproteins demonstrated in this population [21,22].

Acknowledgements This study was performed with the aid of a grant from the Scottish Hospitals Endowment Research Trust (SHERT). The authors would like to thank Gordon Prescott, Statistician, Department of Public Health, University of Aberdeen, for his statistical advice.

References
[1] Ordonez JD, Hiatt RA, Killebrew EJ, Fireman BH. The increased risk of coronary heart disease associated with the nephrotic syndrome. Kidney Int 1993;44:638 42. [2] Warwick GL, Packard CJ. Lipoprotein metabolism in the nephrotic syndrome. Nephrol Dial Transplant 1993;8:385 96. [3] Wheeler DC, Bernard DB. Lipid abnormalities in the nephrotic syndrome: causes, consequences and treatment. Am J Kidney Dis 1994;23:331 46.

[4] Rhadhakrishnan J, Appel A, Valeri A, Appel B. The nephrotic syndrome, lipids and risk factors for cardiovascular disease. Am J Kidney Dis 1993;22:135 42. [5] Joven J, Espinel E, Simo J, Vilella E, Camps J, Oliver A. The inuence of hypoalbuminaemia in the generation of nephrotic hyperlipidaemia. Atherosclerosis 1996;126:243 52. [6] Packard CJ, Shepherd J. Lipoprotein heterogeneity and apolipoprotein B metabolism. Arterioscler Thromb Vasc Biol 1997;17:3542 56. [7] Krauss RM. Heterogeneity of plasma low-density lipoproteins and atherosclerosis risk. Curr Opin Lipidol 1994;5:339 49. [8] Grifn BA, Freeman DJ, Tait GW, et al. Role of plasma triglyceride in the regulation of plasma low density lipoprotein (LDL) subfractions: relative contribution of small, dense LDL to coronary heart disease. Atherosclerosis 1994;106:241 53. [9] Stampfer MJ, Krauss RM, Ma J, et al. A prospective study of triglyceride level, low-density lipoprotein particle diameter, and risk of myocardial infarction. J Am Med Assoc 1996;276:882 8. [10] Deighan CJ, Caslake MJ, McConnell M, Boulton-Jones JM, Packard CJ. Increased atherogenicity of low-density lipoprotein in heavy proteinuria. Nephrol Dial Transplant 1998;13:1183 8. [11] Tomono S, Kawazu S, Kato N, et al. Uptake of remnant like particles (RLP) in diabetic patients from mouse peritoneal macrophages. J Atheroscler Thromb 1994;1:98 102. [12] Masuoka H, Ishikura K, Kamei S, et al. Predictive value of remnant like particles cholesterol/high-density lipoprotein cholesterol ratio as a new indicator of coronary artery disease. Am Heart J 1998;136:226 30. [13] Marcoux C, Tremblay M, Fredenrich A, et al. Plasma remnant like particle and apolipoprotein levels in normolipidaemic and hyperlipidaemic subjects. Atherosclerosis 1998;139:161 71. [14] Lindgren FT, Jensen LC, Hatch FT. The isolation and quantitation analysis of serum lipoproteins. In: Nelson GJ, editor. Blood Lipid and Lipoproteins: Quantitation, Composition and Metabolism. New York: Wiley-Interscience, 1972:181 274. [15] Grifn BA, Caslake MJ, Yip B, Tait GW, Packard CJ, Shepherd J. Rapid isolation of low density lipoprotein (LDL) subfractions from plasma by density gradient ultracentrifugation. Atherosclerosis 1990;83:59 67. [16] Watson TDG, Burns L, Packard CJ, Shepherd J. Selective determination of lipoprotein lipase and hepatic lipase in heparinised plasma from horses. Am J Vet Res 1992;53:771 5. [17] Nakajima K, Saito T, Tamura A, et al. Cholesterol in remnantlike lipoproteins in human serum using monoclonal anti apo B-100 and anti apo A-1 immunoafnity mixed gels. Clin Chim Acta 1993;223:53 71. [18] Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1975;16:31 41. [19] National Cholesterol Education Program. Expert panel on detection, evaluation and treatment of high blood cholesterol in adults. Summary of the second report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation and treatment of high blood cholesterol in adults (ATP II), JAMA,1993, 269, 3015-3023. [20] McNamara JR, Shah PK, Nakajima K, et al. Remnant lipoprotein cholesterol and triglyceride reference ranges from the Framingham Heart Study. Clin Chem 1998;44:1224 32. [21] Warwick GL, Packard CJ, Demant T, Bedford DK, BoultonJones JM, Shepherd J. Metabolism of apolipoprotein B-containing lipoprotein in subjects with nephrotic range proteinuria. Kidney Int 1991;40:129 38. [22] Demant T, Mathes C, Gutlich K, et al. A simultaneous study of the metabolism of apolipoprotein B and albumin in nephrotic patients. Kidney Int 1998;54:2064 80. [23] Tanaka A, Ejiri N, Fujinuma Y, et al. Remnant-like particles and restenosis of coronary arteries after PTCA. Ann NY Acad Sci 1995;748:595 8.

220

C.J. Deighan et al. / Atherosclerosis 157 (2001) 211220 remnant lipoprotein particles. Atherosclerosis 1998;139(2):363 7. [27] Deveraj S, Vega G, Lange R, Grundy SM, Jialal I. Remnant-like particle cholesterol levels in patients with dysbetalipoproteinaemia or coronary artery disease. Am J Med 1998;104:445 50. [28] ONeal D, Lee P, Murphy B, Best J. Low-density lipoprotein particle size distribution in end-stage renal disease treated with haemodialysis or peritoneal dialysis. Am J Kidney Dis 1996;27:84 91. [29] Oda H, Yorioka N, Okushin S, et al. Remnant like particle cholesterol may indicate atherogenic risk in patients on chronic haemodialysis. Nephron 1997;76:7 14.

[24] Takeichi S, Nakajima Y, Osawa M, et al. The possible role of remnant-like particles as a risk factor for sudden cardiac death. Int J Legal Med 1997;110:213 9. [25] Saniabadi AR, Umemura K, Shimoyama M, Adachi M, Nakano M, Nakashima M. Aggregation of human blood platelets by remnant-like lipoprotein particles of plasma chylomicrons and very low density lipoproteins. Thromb Haemost 1997;77:996 1001. [26] Inoue T, Saniabadi AR, Matsunaga R, Hoshi K, Yaguchi I, Morooka S. Impaired endothelial dependent acetylcholine-induced coronary artery relaxation in patients with high serum