Synergistic Effects of APOE and CLU May Increase the Risk of Alzheimer’s Disease: Acceleration of Atrophy in the Volumes and Shapes of the Hippocampus and Amygdala

Abstract

Background:

The volume loss of the hippocampus and amygdala in non-demented individuals has been reported to increase the risk of developing Alzheimer’s disease (AD). Many neuroimaging genetics studies mainly focused on the individual effects of APOE and CLU on neuroimaging to understand their neural mechanisms, whereas their synergistic effects have been rarely studied.

Objective:

To assess whether APOE and CLU have synergetic effects, we investigated the epistatic interaction and combined effects of the two genetic variants on morphological degeneration of hippocampus and amygdala in the non-demented elderly at baseline and 2-year follow-up.

Methods:

Besides the widely-used volume indicator, the surface-based morphometry method was also adopted in this study to evaluate shape alterations.

Results:

Our results showed a synergistic effect of homozygosity for the CLU risk allele C in rs11136000 and APOE ɛ4 on the hippocampal and amygdalar volumes during a 2-year follow-up. Moreover, the combined effects of APOE ɛ4 and CLU C were stronger than either of the individual effects in the atrophy progress of the amygdala.

Conclusion:

These findings indicate that brain morphological changes are caused by more than one gene variant, which may help us to better understand the complex endogenous mechanism of AD.

Keywords

morphometry subcortical structures synergistic

INTRODUCTION

Alzheimer’s disease (AD) is an irreversible neurodegenerative brain disorder caused by genetic and environmental factors [1, 2]. Due to its high mortality [3] and heritability [4] rates, genetic risk factors can serve as powerful markers to identify at-risk individuals for developing AD [5]. Previous reports have shown that gene-gene interactions play a critical role in modulating brain structure and cognitive performance [5 –8]. Information about whether the major risk genes can synergistically accelerate AD-related atrophy of subcortical structures before the onset of dementia symptoms will likely contribute to the identification of at-risk populations [9].

Apolipoprotein E is the major lipid transporter in the brain [10]. It is encoded by the Apolipoprotein E gene (APOE), which has three major polymorphic alleles: ɛ2, ɛ3, and ɛ4 [11] as well as six genotypes: ɛ2/ɛ2, ɛ2/ɛ3, ɛ2/ɛ4, ɛ3/ɛ3, ɛ3/ɛ4, and ɛ4/ɛ4 [12]. Clusterin (CLU) is associated with tissue injury and aging [13], which has two major polymorphic alleles: C and T, three genotypes: CC, CT, and TT. The ɛ4 allele of APOE and the C allele of CLU is thought to increase the risk of developing AD [14, 15]. Numerous studies have found APOE ɛ4 allele effects on entorhinal cortex [16, 17], hippocampal [18 –20], and amygdalar volume [21], as well as temporal lobe volumes [22] in non-demented elderly. These structures are closely related to memory and cognitive performance, supporting them as effective markers during disease progression [23]. Interestingly, the CLU (rs11136000) is associated with worse episodic memory performance in non-demented elders [24]. The hippocampus and amygdala are brain areas crucial for episodic memory, suggesting that the integrity of the hippocampus and amygdala might be injured by CLU [25 –27]. Yet, few studies have investigated the possible effects of the CLU gene on brain structure in non-demented elderly. Recent work has shown that the gene combinations of CLU and APOE additively modified the risk of association with AD [28]. Moreover, previous research has identified that they have additive effects on medial temporal lobe activity [29] and ventricular expansion [30]. APOE and CLU proteins share various major features: they cooperatively suppress amyloid-β (Aβ) deposition [31]; they interact with a shared set of cell-surface receptors [32]; and they promote neurite outgrowth [33, 34]. Because of these biological connections, APOE and CLU polymorphisms may lead to the development of AD by affecting similar pathways [29]. However, to the best of our knowledge, the interaction effects between APOE ɛ4 and CLU C carrier status have been rarely studied [35], and no study has examined the synergistic adverse effects of CLU and APOE on AD-related brain alterations.

The histopathological changes in the early cases of AD are typically seen within the medial temporal lobe beginning in the preclinical phase [36, 37], in particular, the hippocampus and amygdala. Besides, the degeneration of the hippocampus and amygdala in non-demented individuals has been reported to increase the risk of developing AD [38 –40]. Prior works on subcortical structures mainly focused on volumetric methods [5 , 41–44]. Recent studies [45 –47] highlighted the importance of identifying the easily-affected subregions, since the pathological significance of these subregions may be more sensitive. To date, most studies on hippocampal subregions have concentrated on cross-sectional [20 , 49] and a few studies have tried to explore the subregional deformations of the amygdala [48]. Cross-sectional investigations remain inherently limited, as they poorly outline whether abnormalities were dynamically altered over time [50]. In this case, we tried to determine the dynamic changes in subregions of hippocampus and amygdala during 2-year periods, and whether they were detectable in the preclinical phase of AD, which may provide new insights for the comprehension of the effects of multiple risk variants on the hippocampus and amygdala.

The purpose of the current study was to test the hypothesis that APOE ɛ4 and CLU C have synergistic adverse effects on the shapes/volumes of bilateral hippocampi/amygdalae at the baseline and during 2-year periods in non-demented elders. Moreover, we also tested the relations between hippocampus/amygdala atrophy and the gene dose (0, 1, or 2) of APOE ɛ4 or CLU C.

METHODS

Alzheimer’s Disease Neuroimaging Initiative database

Longitudinal data used in this article were collected from the Alzheimer’s disease Neuroimaging Initiative (ADNI) database (http://www.adni.loni.usc.edu). The ADNI, led by Principal InvestigatorMichael W. Weiner MD, was launched in 2003 as apublic-private partnership. ADNI was mainly used to investigate the progression of mild cognitive impairment (MCI) and early AD by using serial magnetic resonance imaging (sMRI), positron emission tomography (PET), and other biological markers. The par-ticipants have been recruited by ADNI from more than 50 sites across the United States and Canada. All participants gave written informed consent. For up-to-date information, visit http://www.adni-info.org.

Participants

The underlying pathology of AD preceded the onset of cognitive symptoms by many years [51]. Thus, censoring risk genetic effects may help early diagnosis and preventative interventions of dementia. Here, we limited the current analyses to normal control (NC) and MCI participants whose genotypes data of CLU at rs11136000 locus and APOE were available. Also, we took into consideration the protective effects of APOE ɛ2 allele against AD, and the APOE ɛ2 carriers were excluded. We included participants with genetic information and four longitudinal MRI scans at baseline, 6-month, 12-month, and 24-month. However, at a time point, if a sample did not obtain MR images or MR images were excluded during data preprocessing, the sample was excluded. Finally, 171 non-demented elders including 82 NC and 89 MCI were selected at all time points. The inclusion criteria for NC individuals were Mini-Mental State Examination (MMSE) scores ranging from 24 to 30 and a Clinical Dementia Rating score of zero. Subjects with MCI had MMSE scores≥24, the Clinical Dementia Rating score of 0.5, and they preserved activities of daily living and the absence of dementia. The details of inclusion and exclusion criteria can be found in [45, 52].

Genetic data

In this article, the APOE and CLU genotypes of participants were obtained from the ADNI database. On the basis of APOE ɛ4 status, the APOE genotype was coded as 0, 1, and 2 for the presence of 0, 1, and 2 APOE ɛ4 allele [53]. Besides, subjects were further coded as 0, 1, and 2, representing the number of CLU C allele, since previous studies have suggested that C allele of CLU rs11136000 was a risk allele for AD [54].

To test the interaction effects, we created 9 gene-gene cohorts based carrying the status of CLU rs11136000 and APOE. According to genotypes of CLU rs11136000, these subjects were divided into 3 groups (38 TT homozygotes; 72 CT heterozygotes; and 61 CC homozygotes). For each of 3 groups, subjects were further divided into ɛ3 homozygotes (ɛ3/ɛ3), ɛ3 heterozygotes (ɛ3/ɛ4), and ɛ4 homozygotes (ɛ4/ɛ4) based on APOE allele status.

To test for combined effects, we created 3 gene-gene cohorts based on the number of the risk alleles in CLU rs11136000 and APOE. Participants with neither APOE ɛ4 allele nor CLU C allele (zero risk allele) were classified into the “0” group (TT+ɛ3/ɛ3); participants with either C allele or ɛ4 allele were classified into the “1” group (TT+ɛ3/ɛ4 or ɛ4/ɛ4; and ɛ3/ɛ3 + CT or CC); participants with more than one risk allele at each locus were classified into the “2” group (ɛ3/ɛ4 + CT or CC; and ɛ4/ɛ4 + CT or CC).

MRI data acquisition

All subjects were scanned with the MR acquisi-tion protocol as previously described in detail [55].In brief, brain MR imaging was acquired at 58 sites, using 1.5T MRI scanners in GE Healthcare, Siemens Medical Solutions USA, or Philips Electronics. High-resolution T1-weighted MRI scans were collected using a sagittal 3-dimensional magnetization-pre-pared rapid gradient echo (3D MP-RAGE) sequence with repetition time = 2400 ms; echo time = 1000 ms; flip angle = 8°. Moreover, image correction proce-dures and post-acquisition correction of certain image artifacts were implemented at a single site to ensure the consistency of these preprocessing steps [53, 55].The volumes of brain structures were processedusing the FreeSurfer version 5.3 software (http://surfer.nmr.mgh.harvard.edu/) based on the 2010 Desikan-Killany atlas [56]. The quality of segmentation of all subcortical structures was manually checked and excluded the images of segmentation errors.

Shape processing

All T1-weighted MR images were automaticallysegmented by FIRST, a model-based subcortical st-ructure integration tool developed as part of FMRIB Software Library (FSL) (https://fmrib.ox.ac.uk/fslwiki/FIRST/). Based on the segmentation of the hippocampus and amygdala, we visually checked the segmented images. Images with segmentation errors were reprocessed and would be excluded if the error could not be corrected. After segmentation for the hippocampus and amygdala, the surface of the hippocampus and amygdala were reconstructed with the topology-preserving level set method [57]. Based on that, the marching cubes algorithm [58] was applied to generate the triangular surface meshes and then the progressive meshes [59] and loop subdivision [60] were used to smooth surfaces and refine the meshes. We also manually checked all the meshes and excluded the images with rough meshes. Subsequently, the conformal grid for each surface was obtained through the holomorphic 1-form basis [61], and then the feature images of the surface were formed by scaling the range of the conformal representation [62]. Finally, the feature images with a chosen template image were aligned via an inverse-consistent surface fluid registration method [62, 63]. More details of the description can be found in [20].

After aligning hippocampus and amygdala surfaces for all subjects, we extracted the radial distance by computing their vertex-wise features. The iso-parametric curve (see red curves in Fig. 1d) is per-pendicular to the medial axis, on the computedconformal grid [64], after which RD value is easily found at each vertex. The radial distance value represents the distance between each surface point and the middle axis [65], as a method to quantify surface deformation. Some studies suggested that the different subregions of the hippocampus and amygdala have different specific functions [66 –68]. To detect the effects of risk-allele on the subdivisions of the hippocampus and amygdala, the radial distance (RD) features were used to generate a distance map in 3D, and smaller RD values were taken as an index of atrophy [69]. Figure 1 gives an overview of processing in this paper.

Fig. 1

The key steps of hippocampus/amygdala morphometry analysis: (a) T1-MR images; (b) MR images were segmented to obtain subcortical structures; (c) The 3D surfaces were reconstructed based on obtained subcortical structures; (d) 3D surface parameterization; (e) Obtaining morphological information; (f) Analyzing the effect of risk variants on the shapes of hippocampus and amygdala.

Statistical associations of the two genotypes with subcortical atrophy

The Hardy-Weinberg equilibrium (HWE) between expected and observed genotypic distributions of both single nucleotide polymorphisms (SNPs) was tested by the HWE version 1.2 [70] program (Col-umbia University, New York). In cross-sectional st-udies, the multivariable linear regression models (MLR) describe the relationship between a set of in-dependent variables and a dependent variable. In thisstudy, the MLR was used to assess the associat-ions between single-gene/interaction/combined effe-cts and morphological changes in the hippocampus and amygdala, after adjusting for baseline age, gender, and total intracranial volume (ICV). Besides, the APOE ɛ4 carrier status was added as a covariate for the main effects model of CLU; likewise, the CLU status was added to covariate for the main effects model of APOE.

The multivariable linear mixed-effects models (MLME) takes into account the group level structure in the data by simultaneously assessing effects within and across groups [71]. We applied the MLME [72 –75] to study the associations between longitudinal change in volumes of hippocampus and amygdala and single-gene/interaction/combined effects at the 2-year follow-up. We denoted the baseline age for the subjects with B_i, the follow-up time for subject i and visit j with X_ij [76], APOE ɛ4 with D_i, CLU with C_i, the ICV with T_i, the interaction with C_iD_i, and the volume Y_ij as the dependent variable. Then the linear mixed model can be constructed by:

$\begin{matrix} Y_{ij} = β_{0} + β_{1} B_{1} + β_{2} X_{ij} + β_{3} T_{i} + β_{4} C_{i} \\ + β_{5} D_{i} + β_{6} C_{i} D_{i} + b_{0 i} + b_{1 i} X_{ij} \end{matrix}$ (1) where β₀,β₁,...,β₆are fixed effects regression coefficients and b_0i, b_1i are random effects regression coefficients [76]. We also considered a simplified model without the interaction term [5, 77]. The variables in both regression models and linear mixed models were converted into a normalized Z-score. In agreement with previous studies [78], we used thep-value to assess the effect degree of combined/int-eraction/main effects. Because a smaller p-value indicated the independent variables had a more significant regression coefficient and more contributor towards the dependent variables. All the statistical analyses were conducted in R v3.6.1 software (The R Foundation).

2.7 Mapping genotype effects across the surfaces of hippocampus and amygdala

The surface features were used to generate distance maps that estimated relationships between the genotypes and regional shape morphometrics of hippocampus and amygdala. These analyses may help to reveal the vulnerable subregions of the hippocampus and amygdala and can observe the trend of def-ormation during 2-year follow-up. Similar to thevolumetric analyses, the MLR and MLME were chosen for main effects analyses of genotype on surface maps at baseline and longitudinal study, respectively. The surface maps indicated that associations between the number of risk alleles (0, 1, or 2) of the genotype and regional shape morphometrics of hippocampus and amygdala. We also examined the epistaticinteractions between APOE status and CLU genotypes on the shape morphometrics of the hippocampus and amygdala.

We thereafter evaluated the combined effects of APOE and CLU on the shape morphometrics of thehippocampus and amygdala. Vertex-based surface maps were corrected for multiple comparisons by thefalse discovery rate (FDR) (q < 0.05) method which was developed by Hochberg and Benjamini [79]. Moreover, baseline age, gender, and ICV were in-cluded as nuisance covariates in all surface-based analyses.

In consideration of the population structures, thesame experiments and statistical analysis were performed. Most of these effects remained significant even after adjusting population structures. The det-ailed information is shown in the Supplementary Material.

RESULTS

Demographic and genetic characteristics

In general, 171 non-demented elderly adults with high-quality imaging data and APOE and CLU gen-otypic information were included in this article. Theobserved distributions of both APOE and CLU genotypes were in HWE (p = 0.422 for APOE and p = 0.063 for CLU). The Chi-square test was used to calculate differences for gender, and the one-wayanalysis of variance (ANOVA) was performed for baseline age, ICV, and MMSE. Among the 3 subgroups (Table 1), there were no significant differences in gender (p = 0.908 for APOE and p = 0.109 forCLU), ICV (p = 0.908 for APOE and p = 0.109 forCLU), MMSE (p = 0.172 for CLU), and age (p = 0.716 for CLU). However, age (p = 0.004) and MMSE (p = 0.004) significantly differed among the APOE subgroups. Post hoc analysis suggested that carrying with ɛ3/ɛ4 (p = 0.024) and ɛ4/ɛ4 (p = 0.002) genotype had lower MMSE score compared with those with ɛ3/ɛ3 genotype. Additionally, subjects with 2 APOE ɛ4 all-eles were younger than those with 0 APOE ɛ4 all-eles (p = 0.002) and those with 1 APOE ɛ4 allele (p = 0.016). The demographic information is shown in Table 1.

Table 1
Demographics characteristics of the genotypic subgroups

Genotypic groups of APOE p Genotypic groups of CLU p

ɛ3/ɛ3 ɛ3/ɛ4 ɛ4/ɛ4 TT CT CC

Age (s) 75.05±5.36 74.1±5.46 69.92±5.75 0.004^* 74.03±7.16 74.75±4.79 74.1±5.29 0.716

Female/Male 42/58 24/35 4/8 0.908 14/24 36/36 20/41 0.109

ICV (cm³) 1561.1±165.4 1559.78±162.6 1524.2±127.5 0.908 1591.69±167.17 1531.54±148.16 1568.36±170.35 0.109

MMSE 28.69±1.58 28.08±1.63 27.17±1.7 0.004^* 28.19±1.53 28.74±1.48 28.05±1.85 0.172

	Genotypic groups of APOE	p	Genotypic groups of CLU	p
Age (s)	75.05±5.36	74.1±5.46	69.92±5.75	0.004^*	74.03±7.16	74.75±4.79	74.1±5.29	0.716
Female/Male	42/58	24/35	4/8	0.908	14/24	36/36	20/41	0.109
ICV (cm³)	1561.1±165.4	1559.78±162.6	1524.2±127.5	0.908	1591.69±167.17	1531.54±148.16	1568.36±170.35	0.109
MMSE	28.69±1.58	28.08±1.63	27.17±1.7	0.004^*	28.19±1.53	28.74±1.48	28.05±1.85	0.172

Data expressed as mean±SD (standard deviation). APOE ɛ4, apolipoprotein E ɛ4; CLU, clusterin protein gene; ICV, total intracranial volume; MMSE, Mini-Mental State Examination. Bold font denotes significant results (p < 0.05).

Volumetric analyses

Main effects of APOE and CLU genotype on hippocampal and amygdalar volumes

To track the atrophy trajectory of hippocampal and amygdalar volumes, we examined whether increasing numbers of risk alleles of APOE or CLU were related to the volume atrophy of bilateral hippocampi and amygdalae. As expected, carrying more APOE ɛ4 alleles was significantly associated with greater atrophy of the hippocampal (left hippocampus: p = 0.029; right hippocampus: p = 0.009, Table 2) and amygdalar (left amygdala: p = 0.018; right amygdala: p = 0.022, Table 2) volumes at baseline; these correlations were enhanced over time as well (Table 2). The main effects of CLU were found in the left amygdala (p = 0.002, Table 3) at baseline. Besides, right amygdalar volume was also affected over time (right amygdala: p = 0.003, Table 3). However, the CLU at rs11136000 locus was not significantly associated with volumes of bilateral hippocampi at the baseline or two-year follow-up (Table 3).

Table 2

Results of multiple linear analyses: Associations between APOE genotype and volume atrophy of hippocampus and amygdala

	Effects of APOE genotype (covariates: Intracranial volume, age, gender, and CLU status)
	Hippocampus	Amygdala
BL
Left hemisphere	β= –0.266	β= –0.280
	p = 0.029	p = 0.018
Right hemisphere	β= –0.309	β= –0.272
	p = 0.009	p = 0.022
24 m
Left hemisphere	β= –0.218	β= –0.252
	p = 8.4×10^-5	p = 1.76×10^-4
Right hemisphere	β= –0.299	β= –0.232
	p = 2.95×10^-6	p = 5.85×10^-4

BL: Baseline; 24 m: at 24-month follow-up. ^a β: standardized coefficient estimate. The negative (–) coefficient estimate signifies that the volume decrease with the number of risk-allele, while the positive (+) coefficient estimate means the increase of volume. ^bp-values are rounded to three decimal places. Bold font denotes significant results (p < 0.05).

Table 3

Results of multiple linear analyses: Associations between CLU genotype and volume atrophy of hippocampus and amygdala

	Effects of CLU genotype (covariates: Intracranial volume, age, gender, and APOE status)
	Hippocampus	Amygdala
BL
Left hemisphere	β = –0.059	β = –0.301
	p = 0.544	p = 0.002
Right hemisphere	β= 0.012	β= –0.272
	p = 0.904	p = 0.077
24 m
Left hemisphere	β= –0.052	β= –0.243
	p = 0.330	p = 9.93×10^-6
Right hemisphere	β= 0.009	β= –0.165
	p = 0.854	p = 0.003

Interactions of APOE and CLU genotypes in the volumes of hippocampus and amygdala

To determine whether APOE variants differently affected hippocampal and amygdalar atrophy in participants with zero, one, or two of CLU C risk all-eles, we introduced the epistatic interaction terms of APOE and CLU in primary models of MLR andMLME containing the main effects of APOE and CLU. The analyses revealed that there were no significant interaction on the hippocampal (left hippocampus: p = 0.328; right hippocampus: p = 0.084, Table 7) and amygdalar (left amygdala: p = 0.118; right amygdala: p = 0.321, Table 7) volumes at baseline. In the follow-up studies, as shown in Fig. 2, the epistatic interactions were present in the left amygdala (p = 0.01, Fig. 2a) and right hippocampus (p = 0.012, Fig. 2b), after controlling for baseline age, gender, and ICV.

Table 7

Results of multiple linear analyses: Interaction effects of APOE status and CLU genotypes on the volume atrophy of hippocampus and amygdala

	Baseline		24-month follow-up
	β	p	β	p
Left_hippocampus	–0.162	0.328	–0.143	0.161
Right_hippocampus	–0.278	0.084	–0.217	0.012
Left_amygdala	–0.251	0.118	–0.234	0.01
Right_amygdala	–0.16	0.321	–0.17	0.063

^a β, standardized coefficient estimate. The negative (–) coefficient estimate signifies that the volume decrease with the number of risk-allele, while the positive (+) coefficient estimate means the increase of volume. ^bp-values are rounded to three decimal places. Bold font denotes significant results (p < 0.05).

Fig. 2

The associations of APOE ɛ4 allele with greater atrophy of left amygdala (a) and right hippocampus (b) in the CLU subgroups at 2-year follow-up.

To further explore the basis of the interactions in volumes of bilateral amygdalae and hippocampi over time, we performed subgroup research. Within the subgroup of CLU CC homozygotes, higher APOE ɛ4 loading was associated with greater atrophy in the volumes of the bilateral amygdalae (Table 4) and hippocampi (Table 5), whereas these associations were not found in the subgroup of CLU TT carriers. Moreover, carrying more ɛ4 allele was connected with only hippocampal volumes in the subgroup of CTheterozygotes (Table 4). All these results demonst-rated that the APOE ɛ4 allele and CLU C homozygosity had a synergistic adverse effect in the volumes of bilateral hippocampi and amygdalae.

Table 4

The results of CLU subgroup research: The associations between APOE ɛ4 genotype and bilateral amygdalae in each subgroup of CLU at 2-year follow-up

	Left_amygdala		Right_amygdala
	β	p	β	p
CLU TT	–1.089	0.522	–0.08	0.617
CLU CT	–0.129	0.222	–1.043	0.306
CLU CC	–0.515	3.33×10^-6	–0.498	1.44×10^-5

Table 5

The results of CLU subgroup research: the associations between APOE ɛ4 genotype and volumes of bilateral hippocampi in each subgroup of CLU at 2-year follow-up

	Left_hippocampus		Right_hippocampus
	β	p	β	p
CLU TT	0.164	0.332	0.001	0.994
CLU CT	–0.380	4.61×10^-4	–0.449	1.47×10^-4
CLU CC	–0.329	0.004	–0.511	3.07×10^-6

Combined effects of APOE and CLU in the volumes of hippocampus and amygdala

Results of the combined effects of APOE and CLU in the volumes of bilateral hippocampi and amygdalae are presented in Table 6. Within the volume analyses, we observed the positive relations between the number of risk alleles and atrophy of bilateral hippocampi/amygdalae. The baseline age, gender, and ICV were included as covariates in MLR and MLME. What is more, these associations became more obvious over time from the longitudinal analysis.

Table 6

Results of multiple linear analyses: Combined effects of APOE status and CLU genotypes on the volume atrophy of hippocampus and amygdala

	Combined effects of APOE and CLU genotypes (covariates: Intracranial volume, age, gender)
	Hippocampus	Amygdala
BL
Left hemisphere	β= –0.173	β= –0.332
	p = 0.115	p = 0.002
Right hemisphere	β= –0.175	β= –0.285
	p = 0.102	p = 0.008
24 m
Left hemisphere	β= –0.132	β= –0.277
	p = 0.025	p = 5.83×10^-6
Right hemisphere	β= – 0.163	β= –0.242
	p = 0.005	p = 7.32×10^-5

Surface-based analyses

3.3.1 Main effects of APOE and CLU genotype on surface morphology

The significant atrophy (cold colors) and expansion (warm colors) subregions were captured by the surface-based statistic maps. The surfaces of the lefthippocampus and left amygdala (shown in the left sides of Fig. 3) were divided into multiple distinct areas based on previous reports [67 , 80]. At baseline, carrying more ɛ4 alleles was associated with greater morphological alterations in the left hippocampus (Fig. 3b) and left amygdala (Fig. 3d). The significant atrophy subregions in the left hippocampus were in Cornu Ammonis (CA) 1 and subiculum, while in left amygdala were located in Amygdalostriate Transition Area (ASTR), Lateral (LA), and Anterior Amygdaloid Area (AAA). Additionally, these relationships also appeared in the subregions of the right hippocampus (Fig. 4a) and right amygdala (Fig. 4c) at 2-year follow-up. The atrophy areas of the right hippocampus were in CA 1 and subiculum, while the atrophy subregions of the left hippocampus (Fig. 4a) were gradually extended from CA 1 and subiculum to CA 2-3 over time. As for the amygdala, the significant atrophy areas of the right amygdala were extensively presented in Basolateral Ventromedial Part (BLVM), Basolateral (BL), ASTR, Central (CE), and AAA subregions, whereas the atrophy subregions of the left amygdala (Fig. 4c) were started in the ASTR, LA, and AAA and extended toward the whole surface gradually over time.

Fig. 3

Surface statistical maps. Top row: (a): the combined effects of APOE and CLU genotypes on shape deformation of the bilateral hippocampi at baseline (uncorrected). (b): the effects of main APOE genotype on shape atrophy of left hippocampus at baseline. Bottom row: (c): the combined effects of APOE and CLU genotypes on shape deformations of bilateral amygdalae at baseline (uncorrected). (d): the effects of main APOE genotype on shape atrophy of left amygdala at baseline. DarkGreen-to-Blue hues indicate subregions in which higher risk-allele loading is associated with atrophy of the surfaces, DarkGreen-to-Red hues indicate regions in which higher risk-allele loading is associated with greater expansion of the surfaces.

Fig. 4

(a): the relationships between carrying more APOE ɛ4 alleles and regional atrophy of the bilateral hippocampi at 2-year follow-up. (b): the combined effects of APOE-CLU on regional deformations of bilateral amygdalae at 2-year follow-up. (c): the effects of APOE genotype on regional atrophy of bilateral amygdalae at 2-year follow-up. (d): the interaction effect of CLU and APOE on regional deformations of the left amygdala at 2-year follow-up. DarkGreen-to-Blue hues indicate subregions in which higher risk-allele loading is associated with atrophy of the surfaces, DarkGreen-to-Red hues indicate regions in which higher risk-allele loading is associated with greater expansion of the surfaces. For all statistical maps, the color bar encodes the FDR-corrected p-values for the observed effects.

However, the effects of CLU genotype on regional atrophy of bilateral hippocampi and amygdalae at baseline and 2-year follow-up after controlling APOE status, baseline age, gender, and ICV, did not pass FDR correction. This may be due to the heavy burden of multiple tests, which were underpowered to detect small genetic effects (15,000 vertexes for each side of the hippocampus).

Interactions of APOE and CLU genotype on surface morphology

APOE status and CLU genotype showed epistatic interactions in regional atrophy of bilateral hippocampi and amygdalae at baseline. However, these influences both disappeared after FDR correction. Moreover, in a 2-year follow-up study, the interactions produced influences on the regional atrophy of bilateral hippocampi and amygdalae, but only the left amygdala (Fig. 4d) survived after FDR correction. The epistatic interaction effects of APOE and CLU on the regional morphological atrophy of the left amygdala were mainly found in the Posterior cortical (PCO), Anterior Cortical (ACO), CE, ASTR, BL, and AAA subregions.

3.3.3 Combined effects of APOE and CLU on surface morphology

The surface-based statistics revealed the combined effects of APOE and CLU genotypes on subregional atrophy of bilateral hippocampi and amygdalae at baseline and 2-year follow-up. At the 2-year follow-up, only bilateral amygdalae (Fig. 4b) remained sig-nificant after FDR correction. At the baseline, thebilateral amygdalae (Fig. 3c) and hippocampi (Fig. 3a) maps did not pass correction for multiple comparisons using an FDR of 5%, so we presented unco-rrected maps. Moreover, we mapped contrast results with uncorrected bilateral hippocampi (Fig. 5) over time. Not all the maps survived correction for multiple comparisons, but interesting trends were still identified. The most atrophy clusters of bilateral amy-gdalae were overlapped at baseline and 2-year follow-up, and the areas of significant atrophy are mainly included BLVM, BL, AAA, and Medial (ME). Bilateral hippocampi also showed atrophy and expansion subregions which were located in the subiculum and CA 1 subregions. Furthermore, the areas of significant atrophy in the bilateral hippocampus mainly located in the subiculum, CA 1, and CA 2-3 subregions, according to the uncorrected maps at 2-year follow-up.

Fig. 5

The uncorrected maps depicting the combined effects of APOE and CLU genotypes on the regional deformations of the bilateral hippocampi at 2-year follow-up. DarkGreen-to-Blue hues indicate subregions in which higher risk-allele loading is associated with atrophy of the surfaces, DarkGreen-to-Red hues indicate regions in which higher risk-allele loading is associated with greater expansion of the surfaces.

DISCUSSION

The main finding of this study is that CLU genotype modulates the effects of APOE ɛ4 on the subcortical structures in both hemispheres at 2-year follow-up. In the subgroup of subjects with C allele homozygotes, with the increasing number of ɛ4 allele, APOE showed additive effects in the volumes of the bilateral amygdalae and hippocampi. The significant APOE-CLU interactions were found in the volumes of the left amygdala and right hippocampus, as well as the shape of the left amygdala; moreover, the combination of APOE ɛ4 allele and CLU C allele showed negative additive effects on the volumes and shapes of bilateral hippocampi and amygdalae. Overall, these results may improve our understanding of the complex roles of multiple genes in the structural architecture of the hippocampus and amygdala.

We observed the subjects with 2 APOE ɛ4 allele were younger than those with 0 or 1 APOE ɛ4 allele. Age was significantly different among the groups and the association between the APOE and hippocampal (or amygdalar) morphometrics may be related toan age effect [48, 81]. To overcome the potential con-founder of age, sex, and brain size, all analyses were corrected for baseline age, sex, and ICV. In agreement with previous studies [82, 83], we found that APOE ɛ4 carriers had inferior cognitive performances as compared to non-ɛ4 carriers in non-demented elders. Previous research demonstrated that the MMSE sco-res were negatively correlated with the volumes of the hippocampus and amygdala [84, 85], suggesting that MMSE score decline was probably a concrete manifestation of morphological atrophy in the hippocampus and amygdala.

Hippocampal and amygdalar degeneration are themost typical characteristic of AD. The volume lossof the hippocampus and amygdala in non-demented individuals has been reported to increase the risk of developing AD [38, 39] and to be associated with higher cerebrospinal fluid tau level (a hallmark of AD pathology) [86 –88]. Recent studies have reported gene dose effects of APOE ɛ4 alleles on the volumes of hippocampus and amygdala in patients with MCI and AD [89, 90], our results indicate that the APOE ɛ4 alleles may affect the hippocampal and amygdala atrophy in a dose-dependent manner in non-demented individuals. In addition, most previous reports [5, 91] have found that the main effects of APOE ɛ4 allele on hippocampal morphology were mainly located in the most vulnerable regions, e.g., CA1 and subiculum [92, 93], our results support these findings and show that this effect is cumulative and spread to the CA 2-3 regions over time. Only one prior research [48] revealed the effect of the APOE ɛ4 genotype on morphological changes of the amygdala. However, they only focused on cross-sectional studies. Our results mapped the morphological deformations in the subregions of the amygdala over time and found the atrophy of surfaces primarily started in the ASTR, LA, and AAA and extended toward the whole surfaces over time. The main effects of CLU were not found on volumes and shapes of the hippocampus, which is in line with previous studies [41, 79]. In contrast, our findings showed that the CLU genotype can exert a significant effect on the volumes of the left amygdala at baseline and bilateral amygdala at 2-year follow-up. Although the mechanisms of the volume loss of the bilateral amygdalae are unclear, we speculate that it may be due to the ventricular expansion affecting the gray and white matter degeneration in nearby subcortical regions [30], for example, the amygdala. However, the effects of CLU variant on regional atrophy of bilateral amygdalae did not pass FDR correction at baseline and 2-year follow-up. This may be due to the heavy burden of multiple tests (15,000 vertexes for each side of the amygdala), which was underpowered to detect small genetic effects [30]. While at least 20 genes have been identified as being associated with AD, APOE is the strongest genetic risk factor [94]. This mechanism also may be consistent with our current observations of volume and shape which show that APOE has a greater effect on hippocampal and amygdala atrophy than CLU.

In the main-effect analysis, we often observed that the hippocampal asymmetry in non-demented elders (R > L), which was primarily due to the greater degree of atrophy of right hippocampal volume with the increasing number of APOE risk-allele. This finding is in line with the results of several other cohort studies [95 –97]. However, for the shape measurements, the atrophic areas of the left hippocampus were more enlarged, after FDR correction. We can only speculate the reason for the inconsistent manifestation of the shape and volume in the asymmetric fashion, which is mainly because of the heavy multiple test burden: the small genetic effects failed to pass the FDR correction. Additionally, among all the analyses, the amygdalar shape was mostly consistent with the amygdalar volume in the asymmetry pattern (L > R). Our results are in agreement with those of a previous study (in terms of APOE genotype’s effects) [98]. These inconsistent findings may reflect the complexity in the effects of genetic factors on brain structures. The same SNP may show a different degree of influence on the different brain structures on the left and right sides.

We also found complex epistatic interaction effects (APOE×CLU) of APOE and CLU on the volumes of the left amygdala and right hippocampus, as well as the shape of left amygdala. In the subgroup of individuals with the CLU CC subgroup, carrying more APOE ɛ4 allele was associated with faster atrophy in bilateral hippocampi and amygdalae at 2-year follow-up. The subjects with both APOE ɛ4 allele and CLU CC show the strongest volume loss of hippocampus and amygdala before the onset of any cognitive symptom, which indicates the synergistic effects of APOE and CLU on the subcortical structures. Numerous studies have reported the APOE ɛ4 allele increases Aβ deposition and reduces Aβ efflux from the brain [99, 100]. Moreover, APOE or CLU knockout in an AD mouse model leads to similar effects on theaccumulation of Aβ, a biomarker of AD [31]. Thus, the APOE and CLU risk genotypes might affect Aβ deposition through similar pathways, ultimately resulting in the development of AD [101]. The potential synergistic effects on Aβ deposition or clearance may have led to the observed APOE-CLU interaction on the vulnerable subcortical structure. Within the subgroup of CLU TT, carrying more APOE ɛ4 allele was not associated with volumes of bilateral hippocampi and amygdalae, whereas higher ɛ4 allele load was significantly correlated with the volume loss in the bilateral hippocampi and amygdalae in the CLU C-allele subgroup. These findings show that CLU might modulate the expression of APOE ɛ4, which may be supported by the findings that CLU and APOE can influence each other’s expression in physiologic and pathologic states [102, 103]. According to the aforementioned findings, we can infer that synergistic effects of APOE and CLU may increase the risk of developing AD by affecting the volume loss in the hippocampus and amygdala at 2-year follow-up.

Interestingly, the combinations of CLU and APOE genotypes have additively modified the risk of AD [28]. Previous research has identified that they have additive effects on medial temporal lobe activity [29] and ventricular expansion [30]. We also examined the combined effects of the two SNPs on the volumes and shapes of the hippocampus and amygdala. Although the CLU risk genetic factor may not individually show different effects on volumes and shapes of the hippocampus, the combination of the APOE and CLU may exhibit meaningful influences on the hippocampus and amygdala. The most obvious observation was that the combination of the two SNPs produced greater reduction in the volumes and shapes of bilateral amygdalae than that of any of the individual SNP at baseline and 2-year follow-up. However, this finding was not replicated in the hippocampal volume. A possible explanation is that the CLU variant may regulate the APOE influences in volumes and shapes of bilateral hippocampi, leading to an insufficient detection power in combined effects. Moreover, the regions of atrophy and expansion regions both presented in the hippocampal surfaces, which may explain why no significant difference between the combined effects and hippocampal volumes at the baseline. The reason for this phenomenon is not yet clear, but we speculate that it may be caused by the artificial grouping of a combination of the two SNPs. Nonetheless, the combination of the two SNPs had stronger effects in the shapes and volumes of the hippocampus and amygdala over time and they were stronger than either of the individual effects in the atrophy progress of the amygdala. Thus, it seems plausible that a combination of APOE ɛ4 allele and CLU C allele may accelerate the course of the disease.

The current study has several potential limitations. First, the number of carriers with two ɛ4 alleles is small. Although the epistatic interactions of APOE and CLU are found in the morphometry of the amygdala and hippocampus at the 2-year follow-up, the sample size of APOE ɛ4 homozygotes carriers may influence statistical ability to detect more subtle eff-ects on brain atrophy. Second, these exploratory obs-ervations need to be confirmed in a larger sample size and longer follow-up. Finally, we only focus on the effects of APOE and CLU, which are only two of multiple AD-related genes. Further work is needed to explore gene-gene interactions among other AD-related genes.

In summary, the present study is the first attempt to explore the epistatic interactions and combined effects between APOE and CLU on the volumes and shape morphometrics of the hippocampus and amygdala in non-demented elders. We found that the APOE ɛ4/ɛ4 and CLU CC have strong synergistic adverse effects on the hippocampal and amygdalar volumes, which indicates that gene-to-gene interaction is a crucial factor for the pathogenesis of AD. The synergistic effects also suggest that subjects with both CLU CC and APOE ɛ4/ɛ4 genotypes are the population who may need more attention in early preventiveinterventions.

Footnotes

ACKNOWLEDGMENTS

This work was supported in part by the National Key Research and Development Program of China (Grant No.2019YFA0706200), in part by the National Natural Science Foundation of China (Grant No.61632014, No.61627808, No.61210010), in part by the National Basic Research Program of China (973 Program, Grant No.2014CB744600), in part by the Gansu Science and Technology Program (Grant No.17JR7WA026), in part by the Program of Beijing Municipal Science & Technology Commission (Grant No.Z171100000117005), in part by the Natural Science Foundation of Gansu Province of China (Grant No.20JR5RA292), and in part by the Fundamental Research Funds for the Central Universities (lzujbky-2018-it67, lzujbky-2018-it64, and lzuxxxy-2018-it70).

Data collection and sharing for this project was funded by the Alzheimer’s Disease NeuroimagingInitiative (ADNI) (National Institutes of Health Gr-ant U01 AG024904) and DOD ADNI (Department of Defense award number W81XWH-12-2-0012). ADNI is funded by the National Institute on Aging, the National Institute of Biomedical Imaging and Bioengineering, and through generous contributions from the following: AbbVie, Alzheimer’s Associat-ion; Alzheimer’s Drug Discovery Foundation; Arac-lon Biotech; BioClinica, Inc.; Biogen; Bristol-Myers Squibb Company; CereSpir, Inc.; Cogstate; Eisai Inc.; Elan Pharmaceuticals, Inc.; Eli Lilly and Company; EuroImmun; F. Hoffmann-La Roche Ltd and its affiliated company Genentech, Inc.; Fujirebio; GE Healthcare; IXICO Ltd.; Janssen Alzheimer Immunotherapy Research & Development, LLC.; Johnson & Johnson Pharmaceutical Research & Development LLC.; Lumosity; Lundbeck; Merck & Co., Inc.; Meso Scale Diagnostics, LLC.; NeuroRx Research; Neurotrack Technologies; Novartis Pharmaceuticals Corporation; Pfizer Inc.; Piramal Imaging; Servier; Takeda Pharmaceutical Company; and Transition Therapeutics. The Canadian Institutes of Health Research is providing funds to support ADNI clinical sites in Canada. Private sector contributions are facilitated by the Foundation for the National Institutes of Health (). The grantee organization is the Northern California Institute for Research and Education, and the study is coordinated by the Alzheimer’s Therapeutic Research Institute at the University of Southern California. ADNI data are disseminated by the Laboratory for Neuro Imaging at the University of Southern California.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Brayne

(1993) Research and Alzheimer’s disease: An epidemiological perspective. Psychol Med 23, 287–296.

Bertram

, Tanzi

(2004) The current status of Alzheimer’s disease genetics: What do we tell the patients? Pharmacol Res 50, 385–396.

Weuve

, Hebert

, Scherr

, Evans

(2014) Deaths in the United States among persons with Alzheimer’s disease (2010–2050). Alzheimers Dement 10, e40–e46.

Gatz

, Reynolds

, Fratiglioni

, Johansson

, Mortimer

, Berg

, Fiske

, Pedersen

(2006) Role of genes and environments for explaining Alzheimer disease. Arch Gen Psychiatry 63, 168–174.

Yang

, Li

, Liu

, Li

, Jiang

(2016) Impact of PICALM and CLU on hippocampal degeneration. Hum Brain Mapp 37, 2419–2430.

Morgen

, Ramirez

, Frölich

, Tost

, Plichta

, Kölsch

, Rakebrandt

, Rienhoff

, Jessen

, Peters

, Jahn

, Luckhaus

, Hüll

, Gertz

H-J

, Schröder

, Hampel

, Teipel

, Pantel

, Heuser

, Wiltfang

, Rüther

, Kornhuber

, Maier

, Meyer-Lindenberg

(2014) Genetic interaction of PICALM and APOE is associated with brain atrophy and cognitive impairment in Alzheimer’s disease. Alzheimers Dement 10, S269–S276.

Zhang

, Qin

, Wang

, Liu

, Zhang

, Jiang

, Yu

(2015) Impacts of PICALM and CLU variants associated with Alzheimer’s disease on the functional connectivity of the hippocampus in healthy young adults. Brain Struct Funct 220, 1463–1475.

Koran

MEI

, Hohman

, Meda

, Thornton-Wells

, Alzheimer’s Disease Neuroimaging Initiative (2014) Genetic interactions within inositol-related pathways are associated with longitudinal changes in ventricle size. J Alzheimers Dis 38, 145–154.

, Lee

, Byun

, Lee

, Ko

, Sohn

, Choe

, Choi

, Baek

, Sohn

C-H

, Kim

, Lee

, KBASE research group (2018) Synergistic interaction between APOE and family history of Alzheimer’s disease on cerebral amyloid deposition and glucose metabolism. Alzheimers Res Ther 10, 1–11.

10.

Nishitsuji

, Hosono

, Nakamura

, Bu

, Michikawa

(2011) Apolipoprotein E regulates the integrity of tight junctions in an isoform-dependent manner in an in vitro blood-brain barrier model. J Biol Chem 286, 17536–17542.

11.

Minta

, Brinkmalm

, Janelidze

, Sjödin

, Portelius

, Stomrud

, Zetterberg

, Blennow

, Hansson

, Andreasson

(2020) Quantification of total apolipoprotein E and its isoforms in cerebrospinal fluid from patients with neurodegenerative diseases. Alzheimers Res Ther 12, 1–11.

12.

Goldberg

, Huey

, Devanand

(2020) Association of APOE e2 genotype with Alzheimer’s and non-Alzheimer’s neurodegenerative pathologies. Nat Commun 11, 1–8.

13.

Yang

C-R

, Leskov

, Hosley-Eberlein

, Criswell

, Pink

, Kinsella

, Boothman

(2000) Nuclear clusterin/XIP8, an x-ray-induced Ku70-binding protein that signals cell death. Proc Natl Acad Sci U S A 97, 5907–5912.

14.

Lambert

J-C

, Heath

, Even

, Campion

, Sleegers

, Hiltunen

, Combarros

, Zelenika

, Bullido

, Tavernier

, Letenneur

, Bettens

, Berr

, Pasquier

, Fiévet

, Barberger-Gateau

, Engelborghs

, Deyn

, Mateo

, Franck

, Helisalmi

, Porcellini

, Hanon

, European Alzheimer’s Disease Initiative Investigators, Pancorbo

MMd

, Lendon

, Dufouil

, Jaillard

, Leveillard

, Alvarez

, Bosco

, Mancuso

, Panza

, Nacmias

, Bossú

, Piccardi

, Annoni

, Seripa

, Galimberti

, Hannequin

, Licastro

, Soininen

, Ritchie

, Blanché

, Dartigues

J-F

, Tzourio

, Gut

, Broeckhoven

, Alpérovitch

, Lathrop

, Amouyel

(2009) Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer’s disease. Nat Genet 41, 1094.

15.

Corder

, Saunders

, Strittmatter

, Schmechel

, Gaskell

, Small

, Roses

, Haines

, Pericak-Vance

(1993) Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families. Science 261, 921–923.

16.

Cardenas

, Chao

, Studholme

, Yaffe

, Miller

, Madison

, Buckley

, Mungas

, Schuff

, Weiner

(2011) Brain atrophy associated with baseline and longitudinal measures of cognition. Neurobiol Aging 32, 572–580.

17.

Bunce

, Anstey

, Cherbuin

, Gautam

, Sachdev

, Easteal

(2012) APOE genotype and entorhinal cortex volume in non-demented community-dwelling adults in midlife and early old age. J Alzheimers Dis 30, 935–942.

18.

Moffat

, Szekely

, Zonderman

, Kabani

, Resnick

(2000) Longitudinal change in hippocampal volume as a function of apolipoprotein E genotype. Neurology 55, 134–136.

19.

Reiman

, Uecker

, Caselli

, Lewis

, Bandy

, Leon

MJd

, Santi

, Convit

, Osborne

, Weaver

, Thibodeau

(1998) Hippocampal volumes in cognitively normal persons at genetic risk for Alzheimer’s disease. Ann Neurol 44, 288–291.

20.

Shi

, Lepore

, Gutman

, Thompson

, Baxter

, Caselli

, Wang

, Alzheimer’s Disease Neuroimaging Initiative (2014) Genetic influence of apolipoprotein E4 genotype on hippocampal morphometry: An N=725 surface-based Alzheimer’s disease neuroimaging initiative study. Hum Brain Mapp 35, 3903–3918.

21.

Honea

, Vidoni

, Harsha

, Burns

(2009) Impact of APOE on the healthy aging brain: A voxel-based MRI and DTI study. J Alzheimers Dis 18, 553–564.

22.

Wishart

, Saykin

, McAllister

, Rabin

, McDonald

, Flashman

, Roth

, Mamourian

, Tsongalis

, Rhodes

(2006) Regional brain atrophy in cognitively intact adults with a single APOE ɛ4 allele. Neurology 67, 1221–1224.

23.

Dong

, Zhang

, Wu

, Li

, Schron

, McMahon

, Shi

, Gutman

, Chen

, Baxter

, Thompson

, Reiman

, Caselli

, Wang

(2019) Applying surface-based hippocampal morphometry to study APOE-E4 allele dose effects in cognitively unimpaired subjects. Neuroimage Clin 22, 101744.

24.

Barral

, Bird

, Goate

, Farlow

, Diaz-Arrastia

, Bennett

, Graff-Radford

, Boeve

, Sweet

, Stern

, Wilson

, Foroud

, Ott

, Mayeux

, National Institute on Aging Late-Onset Alzheimer’s Disease Genetics Study (2012) Genotype patterns at PICALM, CR1, BIN1, CLU, and APOE genes are associated with episodic memory. Neurology 78, 1464–1471.

25.

Pohlack

, Meyer

, Cacciaglia

, Liebscher

, Ridder

, Flor

(2014) Bigger is better! Hippocampal volume and declarative memory performance in healthy young men. Brain Struct Funct 219, 255–267.

26.

Bonner-Jackson

, Mahmoud

, Miller

, Banks

(2015) Verbal and non-verbal memory and hippocampal volumes in a memory clinic population. Alzheimers Res Ther 7, 61.

27.

Markowitsch

, Staniloiu

(2011) Amygdala in action: Relaying biological and social significance to autobiographical memory. Neuropsychologia 49, 718–733.

28.

Santos

LRd

, Almeida

JFF

, Pimassoni

LHS

, Morelato

, Paula

(2020) The combined risk effect among BIN1, CLU, and APOE genes in Alzheimer’s disease. Genet Mol Biol 43, e20180320.

29.

Green

, Gray

, DeYoung

, Mhyre

, Padilla

, DiBattista

, Rebeck

(2014) A combined effect of two Alzheimer’s risk genes on medial temporal activity during executive attention in young adults. Neuropsychologia 56, 1–8.

30.

Roussotte

, Gutman

, Madsen

, Colby

, Thompson

, Alzheimer’s Disease Neuroimaging Initiative (2014) Combined effects of Alzheimer risk variants in the CLU and ApoE genes on ventricular expansion patterns in the elderly. J Neurosci 34, 6537–6545.

31.

DeMattos

, Cirrito

, Parsadanian

, May

, O’Dell

, Taylor

, Harmony

, Aronow

, Bales

, Paul

(2004) ApoE and clusterin cooperatively suppress Aβ levels and deposition: Evidence that ApoE regulates extracellular Aβ metabolism in vivo. Neuron 41, 193–202.

32.

Kounnas

, Loukinova

, Stefansson

, Harmony

, Brewer

, Strickland

, Argraves

(1995) Identification of glycoprotein 330 as an endocytic receptor for apolipoprotein J/clusterin. J Biol Chem 270, 13070–13075.

33.

Nathan

, Bellosta

, Sanan

, Weisgraber

, Mahley

, Pitas

(1994) Differential effects of apolipoproteins E3 and E4 on neuronal growth in vitro. Science 264, 850–852.

34.

Kang

S-W

, Shin

Y-J

, Shim

Y-J

, Jeong

S-Y

, Park

I-S

, Min

B-H

(2005) Clusterin interacts with SCLIP (SCG10-like protein) and promotes neurite outgrowth of PC12 cells. Exp Cell Res 309, 305–315.

35.

Jun

, Naj

, Beecham

, Wang

L-S

, Buros

, Gallins

, Buxbaum

, Ertekin-Taner

, Fallin

, Friedland

, Inzelberg

, Kramer

, Rogaeva

, George-Hyslop

, Arnold

, Baldwin

, Barber

, Beach

, Bigio

, Bird

, Boxer

, Burke

, Cairns

, Carroll

, Chui

, Clark

, Cotman

, Cummings

, DeCarli

, Diaz-Arrastia

, Dick

, Dickson

, Ellis

, Fallon

, Farlow

, Ferris

, Frosch

, Galasko

, Gearing

, Geschwind

, Ghetti

, Gilman

, Giordani

, Glass

, Graff-Radford

, Green

, Growdon

, Hamilton

, Harrell

, Head

, Honig

, Hulette

, Hyman

, Jicha

, Jin

L-W

, Johnson

, Karlawish

, Karydas

, Kaye

, Kim

, Koo

, Kowall

, Lah

, Levey

, Lieberman

, Lopez

, Mack

, Markesbery

, Marson

, Martiniuk

, Masliah

, McKee

, Mesulam

, Miller

, Parisi

, Perl

, Peskind

, Petersen

, Poon

, Quinn

, Murray Raskind , Reisberg

, Ringman

, Roberson

, Rosenberg

, Sano

, Schneider

, Seeley

, Shelanski

, Smith

, Spina

, Stern

, Tanzi

, Trojanowski

, Troncoso

, Deerlin

VMV

, Vinters

, Vonsatte

, Weintraub

, Welsh-Bohmer

, Woltjer

, Younkin

, Cantwell

, Dombroski

, Saykin

, Reiman

, Bennett

, Morris

, Lunetta

, Martin

, Montine

, Goate

, Blacker

, Tsuang

, Beekly

, Cupples

, Hakonarson

, Kukull

, Foroud

, Haines

, Mayeux

, Farrer

, Pericak-Vance

, Schellenberg

, Alzheimer’s Disease Genetics Consortium (2010) Meta-analysis confirms CR1, CLU, and PICALM as Alzheimer disease risk loci and reveals interactions with APOE genotypes. Arch Neurol 67, 1473–1484.

36.

Yang

, Pan

, Song

, Huang

, Li

, Chen

, Gong

, Zhong

, Shi

, Shang

(2012) Voxelwise meta-analysis of gray matter anomalies in Alzheimer’s disease and mild cognitive impairment using anatomic likelihood estimation. J Neurol Sci 316, 21–29.

37.

Scahill

, Schott

, Stevens

, Rossor

, Fox

(2002) Mapping the evolution of regional atrophy in Alzheimer’s disease: Unbiased analysis of fluid-registered serial MRI. Proc Natl Acad Sci U S A 99, 4703–4707.

38.

den Heijer

, van der Lijn

, Koudstaal

, Hofman

, van der Lugt

, Krestin

, Niessen

, Breteler

(2010) A 10-year follow-up of hippocampal volume on magnetic resonance imaging in early dementia and cognitive decline. Brain 133, 1163–1172.

39.

den Heijer

, Geerlings

, Hoebeek

, Hofman

, Koudstaal

, Breteler

(2006) Use of hippocampal and amygdalar volumes on magnetic resonance imaging to predict dementia in cognitively intact elderly people. Arch Gen Psychiatry 63, 57–62.

40.

Tondelli

, Wilcock

, Nichelli

, De Jager

, Jenkinson

, Zamboni

(2012) Structural MRI changes detectable up to ten years before clinical Alzheimer’s disease. Neurobiol Aging 33, 825.e825–825.e836.

41.

Tan

, Wang

H-F

, Tan

M-S

, Tan

C-C

, Zhu

X-C

, Miao

, Yu

W-J

, Jiang

, Tan

, Yu

J-T

, Alzheimer’s Disease Neuroimaging Initiative (2016) Effect of CLU genetic variants on cerebrospinal fluid and neuroimaging markers in healthy, mild cognitive impairment and Alzheimer’s disease cohorts. Sci Rep 6, 26027.

42.

Younes

, Albert

, Miller

, BIOCARD Research Team (2014) Inferring changepoint times of medial temporal lobe morphometric change in preclinical Alzheimer’s disease. Neuroimage Clin 5, 178–187.

43.

Ikram

, Vrooman

, Vernooij

, den Heijer

, Hofman

, Niessen

, van der Lugt

, Koudstaal

, Breteler

(2010) Brain tissue volumes in relation to cognitive function and risk of dementia. Neurobiol Aging 31, 378–386.

44.

Moon

, Lee

, Choi

(2018) Changes in the hippocampal volume and shape in early-onset mild cognitive impairment. Psychiatry Investig 15, 531.

45.

Mattsson

, Andreasson

, Zetterberg

, Blennow

, Alzheimer’s Disease Neuroimaging Initiative (2017) Association of plasma neurofilament light with neurodegeneration in patients with Alzheimer disease. JAMA Neurol 74, 557–566.

46.

Styner

, Lieberman

, Pantazis

, Gerig

(2004) Boundary and medial shape analysis of the hippocampus in schizophrenia. Med Image Anal 8, 197–203.

47.

Shen

, Firpi

, Saykin

, West

(2009) Parametric surface modeling and registration for comparison of manual and automated segmentation of the hippocampus. Hippocampus 19, 588–595.

48.

Tang

, Holland

, Dale

, Miller

, Alzheimer’s Disease Neuroimaging Initiative (2015) APOE affects the volume and shape of the amygdala and the hippocampus in mild cognitive impairment and Alzheimer’s disease: Age matters. J Alzheimers Dis 47, 645–660.

49.

Yao

, Fu

, Wu

, Zhang

, Yu

, Zhang

, Wu

, Wang

, Hu

(2018) Morphological changes in subregions of hippocampus and amygdala in major depressive disorder patients. Brain Imaging Behav 14, 653–667.

50.

Bois

, Levita

, Ripp

, Owens

, Johnstone

, Whalley

, Lawrie

(2016) Longitudinal changes in hippocampal volume in the Edinburgh High Risk Study of Schizophrenia. Schizophr Res 173, 146–151.

51.

, Shi

, Gutman

, Baxter

, Thompson

, Caselli

, Wang

, Alzheimer’s Disease Neuroimaging Initiative (2016) Influence of APOE genotype on hippocampal atrophy over time-an N=1925 surface-based ADNI study. PLoS One 11, e0152901.

52.

Petersen

, Aisen

, Beckett

, Donohue

, Gamst

, Harvey

, Jack

, Jagust

, Shaw

, Toga

, Trojanowski

, Weiner

(2010) Alzheimer’s Disease Neuroimaging Initiative (ADNI): Clinical characterization. Neurology 74, 201–209.

53.

Roussotte

, Gutman

, Madsen

, Colby

, Narr

, Thompson

, the Alzheimer’s Disease Neuroimaging Initiative (2014) The apolipoprotein E epsilon 4 allele is associated with ventricular expansion rate and surface morphology in dementia and normal aging. Neurobiol Aging 35, 1309–1317.

54.

Desikan

, Thompson

, Holland

, Hess

, Brewer

, Zetterberg

, Blennow

, Andreassen

, McEvoy

, Hyman

, Dale

, Alzheimer’s Disease Neuroimaging Initiative Group (2014) The role of clusterin in amyloid-β–associated neurodegeneration. JAMA Neurol 71, 180–187.

55.

Jack

Jr , Bernstein

, Fox

, Thompson

, Alexander

, Harvey

, Borowski

, Britson

PJ, L

, Whitwell

, Ward

, Dale

, Felmlee

, Gunter

, Hill

DLG

, Killiany

, Schuff

, Fox-Bosetti

, Lin

, Studholme

, DeCarli

, Krueger

, Ward

, Metzger

, Scott

, Mallozzi

, Blezek

, Levy

, Debbins

, Fleisher

, Albert

, Green

, Bartzokis

, Glover

, Mugler

, Weiner

(2008) The Alzheimer’s Disease Neuroimaging Initiative (ADNI): MRI methods. J Magn Reson Imaging 27, 685–691.

56.

Desikan

, Ségonne

, Fischl

, Quinn

, Dickerson

, Blacker

, Buckner

, Dale

, Maguire

, Hyman

, Albert

, Killiany

(2006) An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage 31, 968–980.

57.

Han

, Xu

, Prince

(2003) A topology preserving level set method for geometric deformable models. IEEE Trans Pattern Anal Mach Intell 25, 755–768.

58.

Lorensen

, Cline

(1987) Marching cubes: A high resolution 3D surface construction algorithm. ACM Siggraph Computer Graphics 21, 163–169.

59.

Hoppe

Progressive meshes. Proceedings of the 23rd annual conference on Computer graphics and interactive techniques, pp. 99–108.

60.

Loop

(1987) Smooth subdivision surfaces based on triangles.Master’s thesis, University of Utah, Department of Mathematics.

61.

Wang

, Lui

, Gu

, Hayashi

, Chan

, Toga

, Thompson

, Yau

S-T

(2007) Brain surface conformal parameterization using Riemann surface structure. IEEE Trans Med Imaging 26, 853–865.

62.

Shi

, Thompson

, Gutman

, Wang

, Initiative AsDN (2013) Surface fluid registration of conformal representation: Application to detect disease burden and genetic influence on hippocampus. Neuroimage 78, 111–134.

63.

Bro-Nielsen

, Gramkow

(1996) Fast fluid registration of medical images. International Conference on Visualization in Biomedical Computing, pp. 265-276.

64.

Wang

, Song

, Rajagopalan

, An

, Liu

, Chou

Y-Y

, Gutman

, Toga

, Thompson

, Initiative AsDN (2011) Surface-based TBM boosts power to detect disease effects on the brain: An N=804 ADNI study. Neuroimage 56, 1993–2010.

65.

Pizer

, Fritsch

, Yushkevich

, Johnson

, Chaney

(1999) Segmentation, registration, and measurement of shape variation via image object shape. IEEE Trans Med Imaging 18, 851–865.

66.

Maller

, Daskalakis

, Fitzgerald

(2007) Hippocampal volumetrics in depression: The importance of the posterior tail. Hippocampus 17, 1023–1027.

67.

Cavedo

, Boccardi

, Ganzola

, Canu

, Beltramello

, Caltagirone

, Thompson

, Frisoni

(2011) Local amygdala structural differences with 3T MRI in patients with Alzheimer disease. Neurology 76, 727–733.

68.

Ballmaier

, Narr

, Toga

, Elderkin-Thompson

, Thompson

, Hamilton

, Haroon

, Pham

, Heinz

, Kumar

(2008) Hippocampal morphology and distinguishing late-onset from early-onset elderly depression. Am J Psychiatry 165, 229–237.

69.

Thompson

, Hayashi

, de Zubicaray

, Janke

, Rose

, Semple

, Hong

, Herman

, Gravano

, Doddrell

, Toga

(2004) Mapping hippocampal and ventricular change in Alzheimer disease. Neuroimage 22, 1754–1766.

70.

Guo

, Thompson

(1992) Performing the exact test of Hardy-Weinberg proportion for multiple alleles. Biometrics 48, 361–372.

71.

Julián

, Alonso

, Lupón

, Gavidia-Bovadilla

, Ferrer

, de Antonio

, López-Ayerbe

, Domingo

, Santiago-Vacas

, Zamora

, Codina

, Moliner

, Núñez

, Santesmases

, Puig-Domingo

, Bayes-Genis

(2020) Long-term LVEF trajectories in patients with type 2 diabetes and heart failure: Diabetic cardiomyopathy may underlie functional decline. Cardiovasc Diabetol 19, 1–11.

72.

Verbeke

, Molenberghs

(2009) Linear mixed models for longitudinal data, Springer, New York, NY.

73.

Thompson

, Hallmayer

, O’Hara

, Initiative AsDN (2011) Design considerations for characterizing psychiatric trajectories across the lifespan: Application to effects of APOE-ɛ4 on cerebral cortical thickness in Alzheimer’s disease. Am J Psychiatry 168, 894–903.

74.

Fitzmaurice

, Laird

, Ware

Applied longitudinal analysis, John Wiley & Sons.

75.

Bernal-Rusiel

, Greve

, Reuter

, Fischl

, Sabuncu

, Initiative

AsDN

(2013) Statistical analysis of longitudinal neuroimage data with linear mixed effects models. Neuroimage 66, 249–260.

76.

Wachinger

, Salat

, Weiner

, Reuter

, Initiative AsDN (2016) Whole-brain analysis reveals increased neuroanatomical asymmetries in dementia for hippocampus and amygdala. Brain 139, 3253–3266.

77.

Wachinger

, Nho

, Saykin

, Reuter

, Rieckmann

, Alzheimer’s Disease Neuroimaging Initiative (2018) A longitudinal imaging genetics study of neuroanatomical asymmetry in Alzheimer’s disease. Biol Psychiatry 84, 522–530.

78.

Wang

Y-J

, Wan

, Wang

H-F

, Tan

C-C

, Li

J-Q

, Yu

J-T

, Tan

, Alzheimer’s Disease Neuroimaging Initiative (2019) Effects of CD33 variants on neuroimaging biomarkers in non-demented elders. J Alzheimers Dis 68, 757–766.

79.

Biffi

, Anderson

, Desikan

, Sabuncu

, Cortellini

, Schmansky

, Salat

, Rosand

, Alzheimer’s Disease Neuroimaging Initiative (ADNI) (2010) Genetic variation and neuroimaging measures in Alzheimer disease. Arch Neurol 67, 677–685.

80.

Duvernoy

(2005) The human hippocampus: Functional anatomy, vascularization and serial sections with MRI, Springer Science & Business Media.

81.

Crivello

, Lemaître

, Dufouil

, Grassiot

, Delcroix

, Tzourio-Mazoyer

, Tzourio

, Mazoyer

(2010) Effects of ApoE-ɛ4 allele load and age on the rates of grey matter and hippocampal volumes loss in a longitudinal cohort of 1186 healthy elderly persons. Neuroimage 53, 1064–1069.

82.

Quintino-Santos

, Diniz

, Firmo

, Moriguchi

, Lima-Costa

, Castro-Costa

(2015) APOE ɛ4 allele is associated with worse performance in memory dimensions of the mini-mental state examination: The Bambuí Cohort Study of Aging. Int J Geriatr Psychiatry 30, 573–579.

83.

Caselli

, Reiman

, Locke

, Hutton

, Hentz

, Hoffman-Snyder

, Woodruff

, Alexander

, Osborne

(2007) Cognitive domain decline in healthy apolipoprotein E ɛ4 homozygotes before the diagnosis of mild cognitive impairment. Arch Neurol 64, 1306–1311.

84.

Lee

J-Y

, Park

, Mackin

, Ewers

, Chui

, Jagust

, Insel

, Weiner

(2014) Differences in prefrontal, limbic, and white matter lesion volumes according to cognitive status in elderly patients with first-onset subsyndromal depression. PLoS One 9, e87747.

85.

Mullins

, Daly

, Simmons

, Beacher

, Foy

, Lovestone

, Hallahan

, Murphy

(2013) Dementia in Down’s syndrome: An MRI comparison with Alzheimer’s disease in the general population. J Neurodev Disord 5, 19.

86.

De Souza

, Chupin

, Lamari

, Jardel

, Leclercq

, Colliot

, Lehéricy

, Dubois

, Sarazin

(2012) CSF tau markers are correlated with hippocampal volume in Alzheimer’s disease. Neurobiol Aging 33, 1253–1257.

87.

Tsai

C-L

, Liang

C-S

, Lee

J-T

, Su

M-W

, Lin

C-C

, Chu

H-T

, Tsai

C-K

, Lin

G-Y

, Lin

Y-K

, Yang

F-C

(2019) Associations between plasma biomarkers and cognition in patients with Alzheimer’s disease and amnestic mild cognitive impairment: A cross-sectional and longitudinal study. J Clin Med 8, 1893.

88.

Hampel

, Bürger

, Pruessner

, Zinkowski

, DeBernardis

, Kerkman

, Leinsinger

, Evans

, Davies

, Möller

H-J

, Teipel

(2005) Correlation of cerebrospinal fluid levels of tau protein phosphorylated at threonine 231 with rates of hippocampal atrophy in Alzheimer disease. Arch Neurol 62, 770–773.

89.

Hobel

, Isenberg

, Raghupathy

, Mack

, Pa

, Alzheimer’s Disease Neuroimaging Initiative and the Australian Imaging Biomarkers and Lifestyle flagship study of ageing (2019) APOE ɛ4 gene dose and sex effects on Alzheimer’s disease MRI biomarkers in older adults with mild cognitive impairment. J Alzheimers Dis 71, 647–658.

90.

Skouras

, Torner

, Anderson

, Koush

, Falcon

, Minguillon

, Fauria

, Alpiste

, Gispert

, Molinuevo

, the ALFA Study (2019) The effect of APOE genotype and streamline density volume, on hippocampal CA1 down-regulation: A real-time fMRI virtual reality neurofeedback study. bioRxiv, 643577.

91.

Kerchner

, Berdnik

, Shen

, Bernstein

, Fenesy

, Deutsch

, Wyss-Coray

, Rutt

(2014) APOE ɛ4 worsens hippocampal CA1 apical neuropil atrophy and episodic memory. Neurology 82, 691–697.

92.

Pievani

, Galluzzi

, Thompson

, Rasser

, Bonetti

, Frisoni

(2011) APOE4 is associated with greater atrophy of the hippocampal formation in Alzheimer’s disease. Neuroimage 55, 909–919.

93.

Khan

, Westman

, Jones

, Wahlund

L-O

, Mecocci

, Vellas

, Tsolaki

, Kłoszewska

, Soininen

, Spenger

, Lovestone

, Muehlboeck

J-S

, Simmons

, AddNeuroMed consortium and for the Alzheimer’s Disease Neuroimaging Initiative (2015) Automated hippocampal subfield measures as predictors of conversion from mild cognitive impairment to Alzheimer’s disease in two independent cohorts. Brain Topogr 28, 746–759.

94.

Kohannim

, Hua

, Rajagopalan

, Hibar

, Jahanshad

, Grill

, Apostolova

, Toga

, Jack

Jr , Weiner

, Paul Thompson

, Alzheimer’s Disease Neuroimaging Initiative (2013) Multilocus genetic profiling to empower drug trials and predict brain atrophy. Neuroimage Clin 2, 827–835.

95.

, Thompson

, Leow

, Lee

, Yanovsky

, Parikshak

, Khoo

, Wu

, Geschwind

(2011) Apolipoprotein E genotype is associated with temporal and hippocampal atrophy rates in healthy elderly adults: A tensor-based morphometry study. J Alzheimers Dis 23, 433.

96.

O’Dwyer

, Lamberton

, Matura

, Tanner

, Scheibe

, Miller

, Rujescu

, Prvulovic

, Hampel

(2012) Reduced hippocampal volume in healthy young ApoE4 carriers: An MRI study. PLoS One 7, e48895.

97.

Geroldi

, Laakso

, DeCarli

, Beltramello

, Bianchetti

, Soininen

, Trabucchi

, Frisoni

(2000) Apolipoprotein E genotype and hippocampal asymmetry in Alzheimer’s disease: A volumetric MRI study. J Neurol Neurosurg Psychiatry 68, 93–96.

98.

Zhang

, Kong

, Wei

, Zhang

, Ma

, Ba

, Alzheimer’s Disease Neuroimaging Initiative (2020) The effect of ApoE ɛ 4 on clinical and structural MRI markers in prodromal Alzheimer’s disease. Quant Imaging Med Surg 10, 464–474.

99.

Holtzman

(2001) Role of apoE/Aβ interactions in the pathogenesis of Alzheimer’s disease and cerebral amyloid angiopathy. J Mol Neurosci 17, 147–155.

100.

Risacher

, Kim

, Nho

, Foroud

, Shen

, Petersen

, Jack

Jr , Beckett

, Aisen

, Koeppe

, Jagust

, Shaw

, Trojanowski

, Weiner

, Saykin

, Alzheimer’s Disease Neuroimaging Initiative (ADNI) (2015) APOE effect on Alzheimer’s disease biomarkers in older adults with significant memory concern. Alzheimers Dement 11, 1417–1429.

101.

Z-C

, Yu

J-T

, Li

, Tan

(2012) Clusterin in Alzheimer’s disease. Adv Clin Chem 56, 155.

102.

Morrow

, Hatters

, Lu

, Höchtl

, Oberg

, Rupp

, Weisgraber

(2002) Apolipoprotein E4 forms a Molten Globule A potential basis for its association with disease. J Biol Chem 277, 50380–50385.

103.

Wahrle

, Jiang

, Parsadanian

, Legleiter

, Han

, Fryer

, Kowalewski

, Holtzman

(2004) ABCA1 is required for normal central nervous system ApoE levels and for lipidation of astrocyte-secreted apoE. J Biol Chem 279, 40987–40993.

	Genotypic groups of APOE			p	Genotypic groups of CLU			p
	ɛ3/ɛ3	ɛ3/ɛ4	ɛ4/ɛ4		TT	CT	CC
Age (s)	75.05±5.36	74.1±5.46	69.92±5.75	0.004^*	74.03±7.16	74.75±4.79	74.1±5.29	0.716
Female/Male	42/58	24/35	4/8	0.908	14/24	36/36	20/41	0.109
ICV (cm³)	1561.1±165.4	1559.78±162.6	1524.2±127.5	0.908	1591.69±167.17	1531.54±148.16	1568.36±170.35	0.109
MMSE	28.69±1.58	28.08±1.63	27.17±1.7	0.004^*	28.19±1.53	28.74±1.48	28.05±1.85	0.172