This study aims to detect the underlying abnormal functional connectivity strength (FCS) across the whole brain of Apolipoprotein E epsilon 4 allele (APOE-4) carriers.

(1) Subjects and grouping: MR data, APOE-4 information, and other demographic information of all the subjects included in this study was acquired from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database (adni.loni.usc.edu, see http://adni.loni.usc.edu/methods/documents/ for detailed inclusion and exclusion information). Genotyping of APOE (including the three genotypes of APOE ε2, ε3, and ε4) was performed using DNA extracted by Cogenics from a 3 ml aliquot of EDTA blood sample. According to whether carry at least one copy of APOE ε4, all the subjects were divided into APOE4+ group (n = 14, including subjects with APOE3/4 and APOE4/4; male/female, 6/8; age, 71.54 ± 3.96 yrs) and APOE4- group (n = 30, including subjects with APOE2/2, APOE2/3, and APOE3/3; male/female, 14/16; age, 74.22 ± 6.35 yrs). (2) MR acquisition and preprocessing: All the subjects were scanned on 3T Philips MRI scanners. The resting-state functional MR images (rs-fMRI) were acquired using the echo-planar imaging (EPI) sequence with the following parameters: TR = 3000 ms, TE = 30 ms, flip angle = 80°, FOV = 212 mm, voxel size = 3.3×3.3×3.3 mm³, 140 time points. Preprocessing of rs-fMRI data was done using the SPM8 (http://www.fil.ion.ucl.ac.uk/spm/). Major steps included slices timing, head motion correction (six-parameter, excluding > 2.0 mm displacement or > 2.0° angular motion), spatial normalization to the MNI space using DARTEL tool, voxel-wise detrending, and regression of white matter signal, cerebrospinal fluid signal, and six motion parameters. (3) FCS calculation: Firstly, The FCS for a single voxel i was calculated as follow:

$$FCS_{voxel}\left(i\right)=\frac{1}{N}\sum_{j\neq i}r_{ij}$$

Where r_ij represents the Pearson correlation coefficient across the resting-state time series between two voxels of i and j. Negative correlation was excluded due to its biological significance still being unclear, ¹ and a threshold of 0.3 was used to exclude the weak correlations (i.e., r_ij > 0.3 is taken into account). ² Besides, a mask was constructed (thresholding the whole-group mean gray matter probability map with 0.2) to confine the analysis within grey matter voxels. Finally, the resultant FCS maps were performed with a fisher-Z-transform and spatially smoothed with a Gaussian kernel of 8 mm FWHM before group comparison. (4) Statistical analysis: The differences in demographic and clinical information, including age, education level, mini-mental state examination (MMSE) score, were test using the two-sample T-test (or the Mann-Whitney U-test if normal distribution was not given), and a statistical significant threshold of P < 0.05 was used. FCS maps were compared between APOE+ and APOE- groups using a voxel-wise two-sample T-test. Age and MMSE score were added as covariates to regress out their effects. The statistical significance level was set to P < 0.05 (AlphaSim correction).

Age (P = 0.156, T = 1.44), gender (P = 0.813, Χ² =0.056), MMSE score (P = 0.370, Z = 0.896), and education level (P = 0.173, Z = 1.362) were matched between groups. Compared with APOE- group, significantly decreased FCS was found in the bilateral insular, right inferior temporal gyrus, right hippocampus, and right cerebellum in the brain of APOE4+ group (P < 0.05, AlphaSim corrected; Table 1, Fig. 1). Besides, no brain region with significantly increased FCS was found in APOE-4 carriers.APOE-4 is currently known as the strongest genetic risk factor for late-onset Alzheimer’s disease (AD), and its related imaging-genetic studies have aroused increasing scientific interests in recent years. Since AD has been widely accepted as one of the most classic neural disconnection syndrome, identification of its genetic cues at the etiology level by using advanced neuroimaging connectivity probes was attempted in normal population by many studies. ^3-5 Compared with traditional brain connectivity analysis methods where measurement is among and dependent on a series of region of interest (ROI), voxel-wise FSC analysis calculates the connectivity for each voxel across the whole brain, thus provides a comprehensive and spatially detailed representation for the brain connectivity. According to our results, APOE-4 related significant connectivity decrease was mainly distributed in bilateral insular and right temporal gyrus, both of which were considered as significant victim components under AD status. Therefore, future studies are encouraged to further investigate how these differences could be related to the AD generation and progression.This study identifies APOE-4 related significant connectivity decrease at whole-brain level. We hope the finding could help to shed light on the APOE-4’s neuropathological mechanism.