Effect of the amyloidogenic L75P apolipoprotein A-I variant on HDL subpopulations.

BACKGROUND: Hereditary amyloidosis due to mutations of apolipoprotein A-I (apoA-I) is a rare disease characterized by the deposition of amyloid fibrils constituted by the N-terminal fragment of apoA-I in several organs. L75P is a variant of apoA-I associated with systemic amyloidosis predominantly involving the liver, kidneys, and testis, identified in a large number of unrelated subjects. Objective of the present paper was to evaluate the impact of the L75P apoA-I variant on HDL subpopulations and cholesterol esterification in carriers. METHODS AND
RESULTS: Plasma samples were collected from 30 carriers of the amyloidogenic L75P apoA-I (Carriers) and from 15 non affected relatives (Controls). Carriers displayed significantly reduced plasma levels of HDL-cholesterol, apoA-I, and apoA-II compared to Controls. Plasma levels of LpA-I, but not LpA-I:A-II, were significantly reduced in Carriers. HDL subclass distribution was not affected by the presence of the variant. The unesterified to total cholesterol ratio was higher, and cholesterol esterification rate and LCAT activity were lower in Carriers than in Controls.
CONCLUSIONS: The L75P apoA-I variant is associated with hypoalphalipoproteinemia, a selective reduction of LpA-I particles, and a partial defect in cholesterol esterification.

Introduction

Hereditary systemic amyloidoses are autosomal dominant, late-onset disorders caused by mutations in genes coding for a series of plasma proteins, including transthyretin, lysozyme, fibrinogen Aα chain, gelsolin, apolipoprotein A-I (apoA-I) and apoA-II [1]. ApoA-I hereditary amyloidosis (OMIM #107680) is a rare disease characterized by progressive deposition of amyloid fibrils mainly constituted by N-terminal polypeptide fragments of this protein [2]. The clinical manifestations of apoA-I amyloidosis frequently involve liver, kidney, heart, larynx, skin, testis and adrenal glands [2]. The molecular mechanisms underlying the amyloidogenic properties of apoA-I mutants have not been widely investigated. However, it is likely that well-recognized, general mechanisms of in vivo amyloidogenesis, like misfolding of the native protein structure and generation of a critical concentration of the amyloidogenic precursor, may apply also to apoA-I amyloidosis. Indeed, amyloid deposits consisting mainly of N-terminal fragments of the precursor protein are also common in other types of amyloidosis [3].

Fourteen amyloidogenic missense/nonsense apoA-I mutations have been identified thus far, and most of these are private [2,4]. Few years ago we described a new apoA-I variant, L75P, associated with systemic amyloidosis predominantly involving the liver, kidneys, and testis [5]. The variant has been identified in a large number of unrelated subjects and is associated with a milder disease compared with all other known amyloidogenic apoA-I mutations [5]. In the present paper, the impact of the L75P apoA-I variant on high density lipoprotein (HDL) subpopulations and cholesterol esterification was evaluated in carriers in comparison with non carrier relatives.

Methods

Subjects

Thirty heterozygous carriers of the L75P apoA-I variant belonging to 18 Italian families volunteered for this study (Carriers). Fifteen control subjects were selected among non affected relatives (Controls). Carriers were divided into symptomatic and asymptomatic subjects according to the presence or absence of clinical signs of organ damage due to amyloid deposits, evaluated as increased plasma levels of alkaline phosphatase (ALP, liver), of follicle-stimulating hormone and luteinizing hormone (FSH and LH, testis) and as values of glomerular filtration rates (GFR) lower than 60 ml/min/1.73 m2 (kidney). The study was approved by the local institutional Ethic Committee, and all enrolled subjects gave written informed consent for participation in the study.

Lipoprotein profile

Fasting blood was collected into tubes containing Na2-EDTA (final concentration 1 mg/ml) and plasma prepared by low speed centrifugation at 4 °C. Aliquots were immediately frozen and stored at − 80 °C until assayed. Plasma levels of total and unesterified cholesterol, HDL-cholesterol (HDL-C), and triglycerides were determined by certified standard enzymatic techniques. LDL-cholesterol (LDL-C) was calculated using the Friedewald's formula. ApoA-I, apoA-II, and apoB levels were determined by immunoturbidimetry, using commercially available polyclonal antibodies.

Plasma levels of lecithin:cholesterol acyltransferase (LCAT) were measured by a specific competitive ELISA [6]. LCAT ability to esterify cholesterol within endogenous lipoproteins (cholesterol esterification rate, CER) or within a standardized substrate (LCAT activity) was determined as previously described [7].

HDL subpopulations

Plasma levels of HDL particles containing only apoA-I (LpA-I) and of particles containing both apoA-I and apoA-II (LpA-I:A-II) were determined by electroimmunodiffusion in agarose gel [8]. HDL subclass distribution according to particle size was determined by non-denaturing polyacrylamide gradient gel electrophoresis (GGE, 4–30%) of the d < 1.21 g/ml plasma total lipoprotein fraction; the protein-stained gels were scanned with an imaging densitometer to determine particle size and HDL were divided into small (Ø < 8.2 nm), medium (8.2 < Ø<8.8 nm) and large (Ø > 8.8 nm) particles [8]. Plasma content of preβ-HDL was assessed by non-denaturing two-dimensional (2-D) electrophoresis, in which agarose gel electrophoresis was followed by non-denaturing GGE and immunoblotting against apoA-I [9]. Plasma content of preβ-HDL was expressed as percentage of total apoA-I.

Statistical analyses

Results are reported as mean ± SD. Group differences in continuous variables were evaluated by one-way ANOVA, with post hoc analysis by the Neuman–Keuls test. For categorical variables, group differences were examined with the use of 2×2 contingency tables and a χ2 test of significance. Group differences with a p value < 0.05 were considered statistically significant.

Results

Clinical features

Carriers of the L75P apoA-I variant were comparable to Controls for gender distribution and average age (Table 1). Fifteen Carriers were asymptomatic and 15 showed clinical signs of amyloidosis (symptomatic). Asymptomatic Carriers (11 females and 4 males) were prevalently healthy subjects; only one had a previous acute myocardial infarction. Symptomatic Carriers were mainly males (11 out of 15) and had ALP elevation (n = 9), GFR reduction (n = 9) and FSH/LH elevation (n = 9); ten had hypertension, one was obese, one diabetic and one had a previous acute myocardial infarction. On average, symptomatic subjects were older than asymptomatic ones (54.7 ± 11.0 vs 45.4 ± 10.6, respectively, p = 0.026).

Table 1

Lipid and apolipoprotein levels in Carriers and Controls.

	Carriers	Controls	P
Gender (f/m)	15/15	7/8	0.899
Age (y)	50.1 ± 11.6	51.1 ± 11.2	0.784
Total cholesterol (mg/dl)	187.9 ± 42.5	231.7 ± 34.5	0.001
Unesterified cholesterol (mg/dl)	57.4 ± 11.7	63.8 ± 10.2	0.079
Unesterified/total cholesterol	0.31 ± 0.06	0.28 ± 0.04	0.029
LDL-cholesterol (mg/dl)	132.5 ± 39.1	158.6 ± 34.3	0.034
HDL-cholesterol (mg/dl)	33.0 ± 11.1	47.0 ± 9.3	< 0.001
Triglycerides (mg/dl)	109.6 ± 49.6	130.6 ± 44.5	0.174
Apolipoprotein B (mg/dl)	101.9 ± 26.1	110.6 ± 21.4	0.271
Apolipoprotein A-I (mg/dl)	93.7 ± 19.3	117.7 ± 15.1	< 0.001
Apolipoprotein A-II (mg/dl)	29.3 ± 5.1	33.9 ± 5.5	0.008
CER (nmol/ml/h)	27.8 ± 8.0	38.4 ± 13.2	0.002
LCAT activity (nmol/ml/h)	23.8 ± 7.9	41.7 ± 12.1	< 0.001
LCAT (μg/ml)	3.69 ± 1.40	4.24 ± 0.78	0.308

Data are expressed as mean ± SD, except for gender distribution; n = 30 for Carriers and n = 15 for Controls.

According to their clinical features, Carriers were taking hypertensive drugs (n = 9), testosterone replacement therapy (n = 3), aspirin and clopidogrel (n = 2), atorvastatin (n = 2) and insulin (n = 1).

Lipid profile

Carriers displayed significantly lower plasma levels of total and LDL-cholesterol than Controls (Table 1). The L75P apoA-I variant was associated with a significant reduction (− 30%) in HDL-C; apoA-I and apoA-II levels were also significantly reduced in Carriers compared to Controls (Table 1). Interestingly, the plasma unesterified to total cholesterol ratio was higher in Carriers than in Controls (Table 1). The evaluation of the cholesterol esterification process in Carriers showed a significant reduction in cholesterol esterification rate and LCAT activity in the presence of normal plasma LCAT levels (Table 1).

The hypoalphalipoproteinemia observed in Carriers of the L75P apoA-I variant was associated with a marked and significant reduction in plasma LpA-I levels (− 31.5% vs Controls, p < 0.001, Fig. 1), while plasma LpA-I:A-II levels were similar to those observed in Controls. HDL subclass distribution was not affected by the presence of the L75P apoA-I variant; indeed, the percentage of small, medium, and large HDL, and the mean particle diameter of HDL2 and HDL3 were comparable between Carriers and Controls (Table 2). Plasma content of discoidal preβ-HDL tended to be higher in Carriers than in Controls, although the difference was not statistically significant (Table 2). This was associated with an increased unesterified to total cholesterol ratio in the HDL fraction in Carriers compared to Controls (0.27 ± 0.04 vs. 0.20 ± 0.05, P = 0.014).

Fig. 1

Plasma levels of LpA-I (panel A) and LpA-I:A-II (panel B) in Carriers of the L75P apoA-I variant (n = 30) and in Controls (n = 15). Data are expressed as mean ± SD. *p < 0.05 vs. Controls.

Table 2

HDL subpopulations in Carriers and Controls.

	Carriers	Controls	P
Small HDL (%)	14.0 ± 5.5	14.8 ± 5.2	0.642
Medium HDL (%)	17.8 ± 5.0	17.2 ± 4.7	0.701
Large HDL (%)	68.2 ± 9.5	68.0 ± 9.2	0.947
HDL2 (nm)	11.3 ± 0.3	11.4 ± 0.6	0.458
HDL3 (nm)	8.9 ± 0.4	8.9 ± 0.2	1.000
Preβ HDL (%)	19.1 ± 4.3	17.0 ± 4.0	0.122

Data are expressed as mean ± SD; n = 30 for Carriers and n = 15 for Controls.

Impact of disease progression on lipid profile and HDL subclass distribution

Plasma HDL-C levels were nearly significantly reduced, and apoA-I and apoA-II levels tended to be lower in symptomatic compared to asymptomatic Carriers (Table 3). No differences were detected between asymptomatic and symptomatic Carriers in other lipid parameters (Table 3).

Table 3

Lipid/apolipoprotein levels and HDL subpopulations in asymptomatic and symptomatic Carriers.

	Asymptomatic Carriers	Symptomatic Carriers	P
Total cholesterol (mg/dl)	187.0 ± 49.6	188.8 ± 35.7	0.910
Triglycerides (mg/dl)	95.5 ± 50.2	123.7 ± 46.5	0.122
Unesterified cholesterol (mg/dl)	55.7 ± 9.0	59.1 ± 13.9	0.433
Unesterified/total cholesterol	0.31 ± 0.05	0.32 ± 0.06	0.700
LDL-Cholesterol (mg/dl)	130.5 ± 47.1	134.5 ± 30.8	0.785
Apolipoprotein B (mg/dl)	95.4 ± 24.8	108.4 ± 26.6	0.177
HDL-cholesterol (mg/dl)	36.5 ± 12.4	29.5 ± 8.8	0.085
Apolipoprotein A-I (mg/dl)	97.7 ± 22.7	89.7 ± 14.9	0.264
Apolipoprotein A-II (mg/dl)	30.4 ± 6.3	28.4 ± 4.0	0.308
LpA-I (mg/dl)	34.8 ± 7.0	29.7 ± 10.0	0.117
LpA-I:A-II (mg/dl)	62.9 ± 19.5	59.9 ± 13.5	0.628
Small HDL (%)	13.8 ± 6.5	14.3 ± 4.7	0.811
Medium HDL (%)	17.3 ± 5.1	18.3 ± 5.1	0.596
Large HDL (%)	69.0 ± 10.8	67.4 ± 8.3	0.653
HDL2 (nm)	11.3 ± 0.3	11.2 ± 0.3	0.369
HDL3 (nm)	8.9 ± 0.4	9.0 ± 0.4	0.499
Preβ-HDL (%)	18.1 ± 3.7	20.0 ± 4.7	0.229
CER (nmol/ml/h)	34.1 ± 15.2	30.3 ± 19.3	0.554
LCAT activity (nmol/ml/h)	26.2 ± 10.8	22.4 ± 6.2	0.473
LCAT (μg/ml)	4.14 ± 1.38	3.39 ± 1.40	0.151

Data are expressed as mean ± SD; n = 15 in each group.

Discussion

The L75P is an amyloidogenic apoA-I variant associated with systemic amyloidosis predominantly involving the liver, kidneys, and testis [5]. The variant has been identified in a large number of subjects, all living in the same area but not related. Carriers of the L75P apoA-I variant present with a moderate hypoalphalipoproteinemia (30% reduction in HDL-C), which tend to worsen with disease progression. Plasma apoA-I concentrations are also significantly reduced in carriers, similar to what is observed for other amyloidogenic apoA-I mutants. Moreover, in carriers' plasma, the wild-type protein is far more abundant than the mutant, the last accounting for less than 10% of total apoA-I [5,10]. This unbalance, shared by other amyloidogenic apoA-I variants [11,12], is at least in part due to an impaired secretion of the mutant [10], although an enhanced plasma clearance, already reported for the G26R mutant [13], cannot be excluded. The worsening of the hypoalphalipoproteinemia in symptomatic carriers suggests that the low plasma levels of the variant might be also due to a faster uptake in target tissues.

The hypoalphalipoproteinemia observed in carriers of the L75P apoA-I variant is mainly due to a marked reduction in LpA-I concentration with no significant changes in LpA-I:A-II levels. This is somehow surprising, since carriers of apoA-I variants leading to reduced HDL-C levels generally show a decrease of both LpA-I and LpA-I:A-II particles [13-16], or a selective reduction of LpA-I:A-II particles [17,18]. The selective reduction of LpA-I concentrations in carriers could be explained by a disturbed association of the L75P apoA-I with distinct HDL subpopulations (LpA-I particles) or by a preferential hypercatabolism of the variant bound to LpA-I.

A modest but significant increase in the free to total cholesterol plasma ratio and a reduction in cholesterol esterification rate and LCAT activity were detected in carriers of the L75P apoA-I, similar to what is observed in carriers of apoA-I variants with impaired LCAT activation capacity [19,20]. The evaluation of the in vitro capacity of the L75P variant to activate LCAT is far beyond the scope of the present paper; however, it is likely that the L75P apoA-I has impaired LCAT activation ability, despite the mutation being far from the apoA-I central region involved in LCAT activation. Indeed, the majority of the apoA-I mutations causing impaired ability to activate LCAT are located within helices 5, 6, and 7 (residues 121–187) [21]. However, an N-terminal apoA-I mutation (S36A) with impaired in vitro capacity to activate LCAT and associated with impaired cholesterol esterification in carriers has been very recently described [22], suggesting that the N-terminal region of human apoA-I is likely necessary for LCAT activation.

In conclusion, the amyloidogenic L75P apoA-I variant is associated with hypoalphalipoproteinemia, a selective reduction of LpA-I particles, and a partial defect in cholesterol esterification. Despite the hypoalphalipoproteinemia there is no evidence of premature cardiovascular disease in the large kindred; moreover, the evaluation of pre-clinical atherosclerosis, assessed by the measurement of carotid intima-media thickness, showed no differences between carriers and unaffected relatives [23].

Abbreviations

Alkaline phosphatase

apolipoprotein A-I

cholesterol esterification rate

follicle-stimulating hormon

glomerular filtration rate

gradient gel electrophoresis

high density lipoprotein

HDL-cholesterol

LDL-cholesterol

lecithin:cholesterol acyltransferase

luteinizing hormone

particles containing only apoA-I

particles containing both apoA-I and apoA-II