Low-Density Lipoprotein Particle Size Subfractions and Cerebral Amyloidosis

Abstract

Cerebral beta-amyloidosis (CA) is a condition in which amyloid-β (Aβ) proteins are deposited in the cerebral cortex and is a predictor of Alzheimer’s disease (AD). The Aging Brain Study (ABS) investigated risk factors for CA in persons with diabetes and dyslipidemia. In the ABS, we identified that greater levels of LDL cholesterol and lower levels of HDL cholesterol were associated with increased CA. LDL particles comprise multiple species of varying size, density, and protein composition. For example, within a lipoprotein profile characteristic for persons with obesity and diabetic dyslipidemia, larger LDL particles have a greater ApoE to ApoB ratio, enhancing their binding affinity to LDL receptors. The goal of this study was to identify LDL particles that associate with CA in ABS. LDL particle size fractions were measured by ion mobility in plasma samples of 58 participants (40 women and 18 men). CA was assessed using Pittsburgh Compound B index-Positron Emission Tomography (PiB-PET) imaging. Among the LDL subfractions, greater plasma levels of large LDL particles were significantly associated with greater cerebral amyloidosis and lower hippocampal volumes independent of LDL cholesterol or triglyceride levels. Since Aβ is cleared by the LDL receptor family, such as lipoprotein-like receptor 1 (LRP1), one potential mechanism for our findings is competition between ApoE enriched larger LDL particles and brain-derived Aβ on hepatic Aβ clearance and degradation. We conclude that assessing larger LDL particles in persons with atherogenic dyslipidemia may provide a mechanistic biomarker for the extent of CA.

Keywords

Alzheimer’s disease apolipoproteins lipids lipoproteins receptors

INTRODUCTION

Alzheimer’s disease (AD) accounts for 60–80% of dementia cases and is associated with widespread accumulation of amyloid plaques and neurofibrillary tangles in the brain. Cerebral beta-amyloidosis (CA) refers to the deposition of amyloid-β (Aβ) peptide in the cerebral cortex, which precedes the development of neurofibrillary tangles and progressive cognitive decline [1]. Age and presence of the apolipoprotein (APOE) ɛ4 allele are the two strongest risk factors for late-onset AD (LOAD) and associate with greater deposition of Aβ in the brain.

ApoE proteins regulate Aβ clearance [2]. ApoE proteins are required for seeding of brain Aβ plaques [3]. Reducing lipid poor and aggregated ApoE levels using immunotherapies reduces Aβ brain load in mouse models of AD [4, 5]. Aβ is cleared from the brain through the LDL receptor family, most notably lipoprotein related protein-1 (LRP-1) [6], which is located on the abluminal surface of brain capillaries. There is mounting evidence to support that Aβ and ApoE compete for clearance through these receptors, with ApoE4 having a greater binding affinity to LRP-1 than ApoE3 and ApoE2 [2]. LDL receptors are also expressed on liver hepatocytes as well as on endothelial cells lining the blood-brain barrier. In normolipidemic plasma, the majority of ApoE is transported on HDL and very-low density lipoprotein (VLDL) particles, and to lesser extent on LDL particles and in lipid-poor or lipid-free forms.

Ion mobility (IM) is an analytical technique used to quantify lipoprotein particle subfractions that are separated by their size. This method provides accurate, reproducible, and direct determination of size and concentration for a broad range of lipoprotein particles [7]. This method identifies and quantifies large LDL (I, IIa), medium LDL (IIb), small LDL (IIIa) and very small LDL (IIIb, IVa, IVb and IVc) subfractions. Variation in levels and composition of these fractions can have pathophysiologic significance. For example, LDL subfractions in persons with an atherogenic phenotype are characterized by a predominance of small dense LDL, and large LDL fractions (LDL-I) that are enriched with ApoE [8].

In the Aging Brain Study (ABS), we reported a positive association of CA (measured by PiB Index) with fasting levels of LDL cholesterol (LDL-C) and an inverse association with fasting levels of HDL cholesterol (HDL-C) [9, 10]. The goal of the current study was to test whether specific LDL and IDL subfractions measured by IM were associated with amount of CA, as well as measures of hippocampal volumes and cognitive function.

MATERIALS AND METHODS

Participants

Participants were recruited for a multisite research program, the Aging Brain Study, designed to assess vascular contributions to cognitive impairment in individuals with or without AD. We recruited individuals with no or mild cognitive impairment, targeting individuals with substantial vascular disease risk factors, cardiovascular events, and vascular brain injury (VBI) (n = 251) [11]. A subset of participants also underwent ¹¹C-labeled PiB-PET imaging [10] and they are the focus of this study (n = 58).

PET imaging

Cerebral Aβ was measured with PET using the tracer ¹¹C-labeled PiB, which is specifically retained in fibrillar Aβ plaques. The PiB radiotracer was synthesized at Lawrence Berkeley National Laboratory using a previously published protocol. PiB-PET imaging was conducted using a Siemens ECAT HR scanner in 3-D acquisition mode. PiB (10–15 mCi) was injected as a bolus into an antecubital vein after which dynamic acquisition frames were obtained for a total of 90 minutes over progressively longer intervals [12].

Image analysis

PiB data were preprocessed using Statistical Parametric Mapping 8 (SPM8; http://www.fil.ion.ucl.ac.uk/spm/). Frames 6 through 35, as well as the sum of frames 1 to 5, were realigned to frame 17. Realigned frames reflecting the first 20 minutes of acquisition (frames 1–23) were then averaged and used to guide coregistration with the T1-weighted magnetic resonance image (MRI). Distribution volume ratio (DVR) images were generated from PiB frames corresponding to 35–90 min post-injection and quantified using Logan graphical analysis and the participant’s gray matter cerebellar reference region. Cerebral Aβ was quantified using a global PiB index, which averages PiB signal in brain regions with amyloidosis. PET scans were used to measure PiB retention as previously described [12]. DVR values were extracted from regions of interest (ROIs) vulnerable to early Aβ deposition, which include the frontal cortex (anterior to the precentral gyrus), lateral parietal cortex, lateral temporal cortex, posterior cingulate, and precuneus. The occipital cortex was also examined due to its susceptibility to cerebral amyloid angiopathy. ROIs were defined using the Desikan-Killiany atlas and the semi-automated FreeSurfer processing stream. A global measure of PiB uptake (Global PiB Index) was generated in each subject native space by averaging the mean DVR value of these ROIs. This Global PiB Index served as the primary dependent variable. For purposes of describing the sample (but not for purposes of data analysis), PiB positivity indicating significant amyloidosis was defined in eleven young adults (mean age = 24.5, SD = 3.4) who underwent PiB-PET imaging using the same acquisition and processing procedures described above. PiB uptake was determined using DVR values from the Global PiB index. Values 2 SDs above the young average were established as defining values of PiB positivity. Therefore, in the current study, participants with a Global PiB Index ≥1.08 were determined to be PiB-positive.

Measures of brain volume by MRI

Participants underwent MRI using a 3-T scanner (Magnetom Trio System; Siemens) with an 8-channel head coil. Acquired images included a T1-weighted, volumetric, magnetization-prepared rapid gradient-echo (MPRAGE) image (repetition time [TR], 2500 ms; echo time [TE], 2.94 or 2.98 ms; inversion time [T1], 1100 ms) and a fluid-attenuated inversion recovery (FLAIR) image (TR, 5000 ms; TE, 403 ms; TI, 1700 ms; 1.0 × 1.0 mm² in-plane resolution with 1.00 mm thickness). Regional brain areas were normalized to measures of total intracranial volumes.

Neuropsychological testing

All participants received a standardized neuropsychological test battery from which linear measures of global cognition, verbal and non-verbal memory, and executive function were derived using item-response theory, as described previously [13]. Scale development used 400 elderly individuals with cognitive function ranging from normal to demented. Donor items for the global measure came from the first two learning trials of the Memory Assessment Scale (MAS) List Learning Test, Wechsler Memory Scale-Revised Digit Span total raw score, letter fluency (FAS), and animal category fluency. The verbal memory scale combined short delayed free recall, short delayed cued recall, and immediate recall on learning trials 1 and 3 of the MAS List Learning Test. The non-verbal memory scale was derived from items in the Biber serial design learning test. Donor items for executive function were the Initiation-Perseveration subscale of the Mattis Dementia Rating Scale, letter fluency, Digit Span backward, and Spatial Span backward. Each scale was transformed to a mean of 100 and a standard deviation of 15.

Lipoprotein analysis

Particle concentrations of LDL subfractions were analyzed in specific particle-size intervals using IM, which uniquely allows for direct particle quantification as a function of particle diameter (1) following a procedure to remove other plasma proteins (2). The IM instrument utilizes an electrospray to create an aerosol of particles, which then pass through a differential mobility analyzer coupled to a particle counter. Table 1 describes the particle size intervals for the individual LDL subfractions reported in the present study (1). A coefficient of variation (CV) <15% for each subfraction measurement was maintained throughout [7, 14].

Table 1

Lipoprotein Subclass Concentrations using Ion Mobility (unit, nmoL/L)

LDL subfractions		Definition	Size intervals Min and Max (Å)
Very small LDL (vsLDL)	LDL IVc	Low density lipoproteins IVc	180.00	190.00
	LDL IVb	Low density lipoproteins IVb	190.00	199.00
	LDL IVa	Low density lipoproteins IVa	199.00	204.90
	LDL IIIb	Low density lipoproteins IIIb	204.90	208.20
Small LDL (sLDL)	LDL IIIa	Low density lipoproteins IIIa	208.20	214.10
Medium LDL (mLDL)	LDL IIb	Low density lipoproteins IIb	214.10	220.00
Large LDL (lLDL)	LDL IIa	Low density lipoproteins IIa	220.00	224.60
	LDL I	Low density lipoproteins I	224.60	233.30

Statistical analysis

Means (standard deviation) for normally distributed data or median (25th, 75th percentile) for non-normally distributed data were computed. Pearson or Spearman correlation tests were used to correlate the variables. A multivariate model was used to assess the contributions of levels of LDL cholesterol, triglycerides, sex or ApoE genotype on the association of PiB index with LDL size subfractions. The significance was defined by p < 0.05. The data were analyzed using the program R version 3.2.3.

RESULTS

Samples were available from 58 non-demented older participants (40 females and 18 males) who participated in the ABS. Their mean age was 78 years with a range of 67 to 90. Twenty-six participants had a clinical dementia rating (CDR) score of 0, while 32 participants had a score of 0.5. Many participants were overweight with pre- or treated diabetes (Glycosylated hemoglobin, %, mean (SD): 6.6 (1.1)). A subset of the participants had components of atherogenic dyslipidemia: 13 (22%) had elevated triglyceride levels (TG > 150 mg/dL) and 27 (47%) had low HDL cholesterol (HDL-C < 50 mg/dL). Eighteen participants (31%) carried one ApoE4 allele. Additional participant characteristics are summarized in Table 2. All participants underwent PiB PET imaging.

Table 2

Participant Baseline Characteristics

Sex (F/M), number	40/18
Age, y	77.8 (6.3)
Education level, y	14.4 (2.6)
Blood pressure, mm Hg
SBP	142.3 (17.8)
DBP	69.6.6 (10.3)
BMI, kg/m²	28.4 (5.6)
Lipids, mg/dL
Total Cholesterol	171.8 (41.7)
LDL Cholesterol	90.1 (30.1)
HDL Cholesterol	54.8 (154.3)
Triglyceride^*	97 (62)
Glycosylated Hemoglobin, %	6.6 (1.1)
Fasting Glucose, mg/dL	116.9 (40.1)
hsCRP (mg/L)^*	2.6 (2.1)

Means (standard deviation) for normally distributed data, or ^*median (25th, 75th percentile) for non-normally distributed data are presented.

As expected, levels of LDL particle subfractions were significantly correlated with levels of LDL-C (Table 3), with the strongest correlation observed between large LDL particles and LDL-C. Levels of triglyceride as well as glycosylated hemoglobin were correlated with concentrations of very small LDL particles but not with other LDL subfractions. No correlations were observed between levels of LDL subfractions and HDL cholesterol.

Table 3

Correlation of biomarkers with LDL subfraction particle concentrations (nmoL/L)

	Large LDL	Medium LDL	Small LDL	Very small LDL
LDL-C (mg/dL)	r = 0.67, p < 0.001	r = 0.41, p = 0.001	r = 0.38, p = 0.003	r = 0.32, p = 0.005
HDL-C (mg/dL)	r= –0.04, p = 0.71	r= 0.03, p = 0.97	r= 0.07, p = 0.6	r= 0.13, p = 0.3
Triglyceride (mg/dL)^*	r= 0.04, p = 0.75	r= –0.21, p = 0.12	r= 0.01, p = 0.99	r = 0.2, p = 0.003
Glycosylated hemoglobin (%)	r= –0.06, p = 0.65	r= 0.01, p = 0.92	r= 0.21, p = 0.15	r = 0.32, p = 0.01
hsCRP (mg/L)^*	r= 0.06, p = 0.65	r= –0.9, p = 0.52	r= 0.04, p = 0.73	r= 0.07, p = 0.62

Pearson or Spearman correlation(^*) coefficients with p values are presented. Significant correlations are in bold.

Levels of large and medium sized LDL particles were found to be most strongly correlated with PiB index (large LDL: r= 0.47, p < 0.001, Fig. 1A, medium LDL, r= 0.51, p < 0.001, Fig. 1B). The association of large LDL with PiB index remained significant after removing one outlier value (excluding one sample where large LDL > 400 nmol/L, r= 0.38, p = 0.003). The relationship between small LDL particles and PiB was significant but weaker (r= 0.42, p = 0.01, Fig. 1C). Very small LDL particles did not correlate with PiB index (r= 0.14, p = 0.27, Fig. 1D). Using a multivariate model, the associations of large and medium sized LDL particles with PiB index were independent of LDL-C or TG levels. The levels of LDL subfractions did not significantly differ by ApoE4 status or by sex.

Fig.1

Association of LDL subfractions with cerebral amyloidosis (PiB index). Greater levels of large LDL and medium LDL particles were significantly associated with greater cerebral amyloidosis. (large LDL: r= 0.47, p < 0.001, A; medium LDL, r= 0.51, p < 0.001, B; small LDL particles: r= 0.42, p = 0.01, C; very small LDL: r= 0.14, p = 0.27, D).

Hippocampal volume assessed by MRI brain imaging after adjusting for intracranial volume was significantly inversely correlated with large LDL particle levels (r= –0.32, p = 0.02, Fig. 2A), but the association with the other LDL subfractions did not reach statistical significance (Fig. 2B–D). No significant association was observed between the levels of any LDL subfraction and measures of cognition (data not shown) as measured by neuropsychological testing.

Fig.2

Association of LDL subfractions with hippocampal volumes. Greater levels of large LDL particles were significantly associated with lower hippocampal volumes (r= –0.32, p = 0.02, A). The association with the other LDL subfractions did not reach statistical significance (medium LDL, r= –0.21, p = 0.13, B; small LDL, r= –0.23, p = 0.1, C; very small LDL, r= –0.17, p = 0.21, D). The hippocampal volume is normalized to the intracranial volume (ICV).

DISCUSSION

Epidemiological studies have found that high cholesterol levels are correlated with increased AD dementia risk [15, 16]. Furthermore, analysis of two large independent genome-wide association studies of LOAD strongly implicate genetic variation in lipid metabolism as a cause of LOAD susceptibility [17]. However, the mechanisms for effects of cholesterol metabolism on AD risk are not well understood. The ABS was designed to investigate mechanisms of increased CA in persons with diabetes and dyslipidemia. In this study, increased LDL-C levels were associated with increased CA [18]. The present report provides the first detailed characterization of associations between levels of specific LDL particle subfractions with both CA and hippocampal volume in ABS. Our findings demonstrate that in this population, the levels of larger LDL particles were significantly associated with greater CA and lower hippocampal volumes, two correlates of AD.

In patients with obesity and hypertriglyceridemia, the liver secretes a higher percentage of ApoE and ApoC-III enriched VLDL, IDL, and LDL particles [19], and there is a predominance of small dense LDL particles [20]. As VLDL is lipolyzed by lipoprotein lipase into intermediate density lipoprotein (IDL) and LDL, the large LDL particles in this dyslipidemia acquire a higher ApoE to ApoB ratio [8]. It is estimated that in these patients, there is, on average, 1 molecule of ApoE for every 6 LDL-I particles, whereas in unaffected controls, the ratio is 1 to 17 [8]. The greater ApoE to ApoB ratio can confer greater binding affinity to LDL receptors such as LRP-1 [8]. LRP-1 participates in Aβ clearance from the liver. There is evidence to support that ApoE competes with Aβ for its clearance [2]. Therefore, one potential mechanism for our findings is competition between plasma-derived ApoE-containing LDL particles and brain Aβ that may retard hepatic Aβ clearance and degradation as illustrated in Fig. 3.

Fig.3

A model for a potential mechanism for the association of larger LDL particles with CA. VLDLs are catabolized by LPL to IDLs, and a large proportion of ApoE proteins are transferred to HDLs in the process of LDL formation. It is estimated that for every 17 LDL particles, one large LDL particle contains a molecule of ApoE (A). In atherogenic dyslipidemia, lipolysis of an increased number of ApoE enriched VLDL particles results in a greater percentage of large LDL particles that contain ApoE. It is estimated that for every 6 of these LDL particles, one contains a molecule of ApoE. Since ApoE-containing large LDLs have high affinity for LDL receptors, we hypothesize that these particles compete with Aβ for hepatic clearance. LPL, lipoprotein lipase; VLDL, very low density lipoprotein; IDL, intermediate density lipoprotein; L-LDL, large LDL; HDL, high density lipoprotein.

It is interesting to note that the ApoE4 genotype, which is a major risk factor for CA, is also associated with a greater number of ApoE molecules per HDL-like particle in the brain [21]. ApoE4 has a greater ability to bind lipids than ApoE3 [22]. When added to plasma, ApoE3 binds preferentially to HDL and ApoE4 binds more to VLDL [23]. ApoE4’s greater ability to bind VLDL impairs lipoprotein lipase mediated lipolysis [24], a process that regulates LDL particle size, and may explain the increase in LDL particle size in subjects with the ApoE 4/3 genotype compared with those with ApoE 3/3 and ApoE 3/2 [25]. ApoE4 containing lipoprotein particles have greater affinity to LDL receptors compared with those containing ApoE2 [26]. We hypothesize that conditions favoring an increase in the affinity of ApoE to LDL receptors are important mechanisms for CA by affecting hepatic Aβ clearance and degradation through the LDL receptor family.

We believe that our findings are of importance for several reasons. First, the association of larger LDL subfractions with CA was independent of LDL-C and TG levels, highlighting the value for assessing LDL particle size to better define AD risk factors. Second, identifying these particles provides functional insight into the mechanisms for CA as described above. Third, assessing LDL particle size provides biomarkers for interventions that can be tailored to selectively lower the larger LDL particles. The possible benefits of treatment directed toward lower specific LDL subfractions enriched in ApoE could therefore modify the risk of AD. We believe the next step is to better under the mechanisms through which lipoproteins modulate Aβ, providing new approaches to slowing Aβ deposition and thus potentially reduce the incidence of AD.

Our study has several limitations. First, plasma sample volumes were not sufficient to assess the apolipoprotein content of the lipoprotein subfractions. In addition, the sample size was small, limiting our ability to detect differences by ApoE genotype or by sex. Third, there were significantly more females than males, but the study was not powered to detect a sex difference. Finally, the analyses were cross sectional and a longitudinal data would be needed to establish the capacity of larger LDL particle concentrations to predict cognitive decline and AD risk.

In conclusion, this study identifies larger LDL particles in persons with atherogenic dyslipidemia as being independently associated with CA. The mechanisms whereby these particles may enhance CA merit additional investigation.

Footnotes

ACKNOWLEDGMENTS

Dr. Yassine was supported by R21AG056518, R01AG055770, and R01AG054434 from the National Institute of Aging. This work was also supported by grants P01AG12435 (HC, BR,), AG031563 (BR), and P50AG05142 (HC) from the National Institutes of Health. We wish to thank Ashley Martinez for creating the illustration. This article was prepared while Dr. Reed was employed at University of California, Davis. The opinions expressed in this article are the author’s own and do not reflect the view of the National Institutes of Health, the Department of Health and Human Services, or the United States government.

Authors’ disclosures available online ().

References

Eisele

, Fritschi

, Hamaguchi

, Obermüller

, Füger

, Skodras

, Schäfer

, Odenthal

, Heikenwalder

, Staufenbiel

, Jucker

(2014) Multiple factors contribute to the peripheral induction of cerebral β-amyloidosis. J Neurosci 34, 10264–10273.

Kanekiyo

, Xu

, Bu

(2014) ApoE and Aβ in Alzheimer’s disease: Accidental encounters or partners? Neuron 81, 740–754.

Liu

, Zhao

, Fu

, Wang

, Linares

, Tsai

, Bu

(2017) ApoE4 accelerates early seeding of amyloid pathology. Neuron 96, 1024–1032.e1023.

Liao

, Hori

, Hudry

, Bauer

, Jiang

, Mahan

, Lefton

, Zhang

, Dearborn

, Kim

, Culver

, Betensky

, Wozniak

, Hyman

, Holtzman

(2014) Anti-ApoE antibody given after plaque onset decreases Abeta accumulation and improves brain function in a mouse model of Abeta amyloidosis. J Neurosci 34, 7281–7292.

Liao

, Li

, Xiong

, Bien-Ly

, Jiang

, Zhang

, Finn

, Hoyle

, Keyser

, Lefton

, Robinson

, Serrano

, Silverman

, Guo

, Getz

, Henne

, Leyns

, Gallardo

, Ulrich

, Sullivan

, Lerner

, Hudry

, Sweeney

, Dennis

, Hyman

, Watts

, Holtzman

(2018) Targeting of nonlipidated, aggregated apoE with antibodies inhibits amyloid accumulation. J Clin Invest 128, 2144–2155.

Andersen

, Willnow

(2006) Lipoprotein receptors in Alzheimer’s disease. Trends Neurosci 29, 687–694.

Caulfield

, Li

, Lee

, Blanche

, Salameh

, Benner

, Reitz

, Krauss

(2008) Direct determination of lipoprotein particle sizes and concentrations by ion mobility analysis. Clin Chem 54, 1307–1316.

Barbagallo

, Levine

, Blanche

, Ishida

, Krauss

(1998) Influence of apoE content on receptor binding of large, buoyant LDL in subjects with different LDL subclass phenotypes. Arterioscler Thromb Vasc Biol 18, 466–472.

Reed

, Marchant

, Jagust

, DeCarli

, Mack

, Chui

(2012) Coronary risk correlates with cerebral amyloid deposition. Neurobiol Aging 33, 1979–1987.

10.

Reed

, Villeneuve

, Mack

, DeCarli

, Chui

, Jagust

(2014) Associations between serum cholesterol levels and cerebral amyloidosis. JAMA Neurol 71, 195–200.

11.

Frazier

, Seider

, Bettcher

, Mack

, Jastrzab

, Chao

, Weiner

, DeCarli

, Reed

, Mungas

, Chui

, Kramer

(2014) The role of carotid intima-media thickness in predicting longitudinal cognitive function in an older adult cohort. Cerebrovasc Dis 38, 441–447.

12.

Marchant

, Reed

, Sanossian

, Madison

, Kriger

, Dhada

, Mack

, DeCarli

, Weiner

, Mungas

, Chui

, Jagust

(2013) The aging brain and cognition: Contribution of vascular injury and abeta to mild cognitive dysfunction. JAMA Neurol 70, 488–495.

13.

Reed

, Mungas

, Kramer

, Ellis

, Vinters

, Zarow

, Jagust

, Chui

(2007) Profiles of neuropsychological impairment in autopsy-defined Alzheimer’s disease and cerebrovascular disease. Brain 130, 731–739.

14.

Mora

, Caulfield

, Wohlgemuth

, Chen

, Superko

, Rowland

, Glynn

, Ridker

, Krauss

(2015) Atherogenic lipoprotein subfractions determined by ion mobility and first cardiovascular events after random allocation to high-intensity statin or placebo: The Justification for the Use of Statins in Prevention: An Intervention Trial Evaluating Rosuvastatin (JUPITER) Trial. Circulation 132, 2220–2229.

15.

Solomon

, Kivipelto

, Wolozin

, Zhou

, Whitmer

(2009) Midlife serum cholesterol and increased risk of Alzheimer’s and vascular dementia three decades later. Dement Geriatr Cogn Disord 28, 75–80.

16.

Notkola

, Sulkava

, Pekkanen

, Erkinjuntti

, Ehnholm

, Kivinen

, Tuomilehto

, Nissinen

(1998) Serum total cholesterol, apolipoprotein E epsilon 4 allele, and Alzheimer’s disease. Neuroepidemiology 17, 14–20.

17.

Jones

, Holmans

, Hamshere

, Harold

, Moskvina

, Ivanov

, Pocklington

, Abraham

, Hollingworth

, Sims

, Gerrish

, Pahwa

, Jones

, Stretton

, Morgan

, Lovestone

, Powell

, Proitsi

, Lupton

, Brayne

, Rubinsztein

, Gill

, Lawlor

, Lynch

, Morgan

, Brown

, Passmore

, Craig

, McGuinness

, Todd

, Holmes

, Mann

, Smith

, Love

, Kehoe

, Mead

, Fox

, Rossor

, Collinge

, Maier

, Jessen

, Schürmann

, Heun

, Kölsch

, van den Bussche

, Heuser

, Peters

, Kornhuber

, Wiltfang

, Dichgans

, Frölich

, Hampel

, Hüll

, Rujescu

, Goate

, Kauwe

, Cruchaga

, Nowotny

, Morris

, Mayo

, Livingston

, Bass

, Gurling

, McQuillin

, Gwilliam

, Deloukas

, Al-Chalabi

, Shaw

, Singleton

, Guerreiro

, Mühleisen

, Nöthen

, Moebus

, Jöckel

, Klopp

, Wichmann

, Rüther

, Carrasquillo

, Pankratz

, Younkin

, Hardy

, O’Donovan

, Owen

, Williams

(2010) Genetic evidence implicates the immune system and cholesterol metabolism in the aetiology of Alzheimer’s disease. PloS One 5, e13950.

18.

Reed B

, Mack

, DeCarli

, Chui

, Jagust

(2014) Associations between serum cholesterol levels and cerebral amyloidosis. JAMA Neurol 71, 195–200.

19.

Zheng

, Khoo

, Ikewaki

, Sacks

(2007) Rapid turnover of apolipoprotein C-III-containing triglyceride-rich lipoproteins contributing to the formation of LDL subfractions. J Lipid Res 48, 1190–1203.

20.

Siri-Tarino

, Williams

, Fernstrom

, Rawlings

, Krauss

(2009) Reversal of small, dense LDL subclass phenotype by normalization of adiposity. Obesity 17, 1768–1775.

21.

Gong

, Kobayashi

, Hayashi

, Zou

, Sawamura

, Fujita

, Yanagisawa

, Michikawa

(2002) Apolipoprotein E (ApoE) isoform-dependent lipid release from astrocytes prepared from human ApoE3 and ApoE4 knock-in mice. J Biol Chem 277, 29919–29926.

22.

Nguyen

, Dhanasekaran

, Nickel

, Nakatani

, Saito

, Phillips

, Lund-Katz

(2010) Molecular basis for the differences in lipid and lipoprotein binding properties of human apolipoproteins E3 and E4. Biochemistry 49, 10881–10889.

23.

Weisgraber

(1990) Apolipoprotein E distribution among human plasma lipoproteins: Role of the cysteine-arginine interchange at residue 112. J Lipid Res 31, 1503–1511.

24.

, Dhanasekaran

, Alexander

, Rader

, Phillips

, Lund-Katz

(2013) Molecular mechanisms responsible for the differential effects of apoE3 and apoE4 on plasma lipoprotein-cholesterol levels. Arterioscler Thromb Vasc Biol 33, 687–693.

25.

Moreno

, Pérez-Jiménez

, Marín

, Gómez

, Pérez-Martínez

, Moreno

, Bellido

, Fuentes

, López-Miranda

(2004) The effect of dietary fat on LDL size is influenced by apolipoprotein E genotype in healthy subjects. J Nutr 134, 2517–2522.

26.

Weisgraber

, Innerarity

, Mahley

(1982) Abnormal lipoprotein receptor-binding activity of the human E apoprotein due to cysteine-arginine interchange at a single site. J Biol Chem 257, 2518–2521.