APOE4 and age affect the brain entorhinal cortex structure and blood arachidonic acid and docosahexaenoic acid levels after mild TBI

Abstract

A reduction in the thickness and volume of the brain entorhinal cortex (EC), together with changes in blood arachidonic acid (AA) and docosahexaenoic acid (DHA), are associated with Alzheimer's disease (AD) among apolipoprotein E ε4 carriers. Magnetic Resonance Imaging (n = 631) and plasma lipidomics (n = 181) were performed using the LIMBIC/CENC cohort to examine the influence of ε4 on AA- and DHA-lipids and EC thickness and volume in relation to mild traumatic brain injury (mTBI). Results showed that left EC thickness was higher among ε4 carriers with mTBI. Repeated mTBI (r-mTBI) was associated with reduced right EC thickness after controlling for ε4, age and sex. Age, plus mTBI chronicity were linked to increased EC White Matter Volume (WMV). After controlling for age and sex, the advancing age of ε4 carriers with blast mTBI was associated with reduced right EC Grey Matter Volume (GMV) and thickness. Among ε4 carriers, plasma tau and Aβ40 were associated with mTBI and blast mTBI, respectively. Chronic mTBI, ε4 and AA to DHA ratios in phosphatidylcholine, ethanolamides, and phosphatidylethanolamine were associated with decreased left EC GMV and WMV. Further research is needed to explore these as biomarkers for detecting AD pathology following mTBI.

Keywords: Apolipoprotein E; Arachidonic acid; Blast injury; Docosahexaenoic acid; Entorhinal cortex; Repetitive mild traumatic brain injury; Traumatic brain injury.

Fig. 1

Influence of the ε4 allele on mTBI and blast injuries on thickness and volume of the EC. (A-D) Means ± SE (A) Left EC thickness stratified by the presence of the ε4 allele and mTBI diagnosis (ε4- control = 95, ε4 + controls = 33, ε4- mTBI = 375 and ε4 + mTBI = 128). Among controls, the thickness of the left EC was smaller among ε4 carriers compared to non-ε4 carriers. Compared to ε4 controls, ε4 + mTBI patients had higher left EC thickness. (B) There were no significant differences between mTBI patients and controls and no influence of the ε4 status on the right EC thickness. (C) Left GMW stratified by the presence of the ε4 allele and blast mTBI (ε4- non-blast = 318, ε4 + non-blast (including healthy controls and those without blast-related mTBI) = 106, ε4- blast = 183, ε4 + blast = 60). There were no differences between the groups. (D) Compared to the ε4 + non-blast group, ε4 + patients with blast mTBI had higher GMV in the right EC. *p < 0.05 for ANOVA followed by B-H correction. (E–F) Heatmaps showing beta-coefficients from linear regression analyses. (E) The interaction between the presence of the ε4 allele and an increase in the number of blast mTBI per participant was associated with an increase in the GMV of the left and the right EC. However, interactions between age, the presence of the ε4 allele and the numbers of blast mTBI were negatively associated with lowering of the right EC GMW and thickness. (F) Interaction between the presence of the ε4 allele and years since the 1^st mTBI was associated with increases of the WMV in the right and the left EC, whereas the interactions between the presence of the ε4 allele, age and years since the 1^st mTBI is the associated with higher WMV in the left EC and lower WMV in the right EC. All regression analyses were adjusted for the independent effects of the ε4 allele, age and sex.

Fig. 2

Association of protein and lipid biomarkers with mTBI diagnosis. Each image is a heat map of odds ratios and corresponding p-values for each protein and AA and DHA ratios of each lipid class. (A) Logistic regression controlling for the independent effects of age, sex, and ε4 as well as incorporating interactive effects of each protein (tau, Aβ40, Aβ42, Aβ40/Aβ42 ratios and NfL) with ε4 and with ε4 plus age were performed to calculate the OR for estimating their association with a diagnose of mTBI. While plasma total tau in the presence of the ε4 allele is associated with a mTBI diagnosis. The OR for the mTBI diagnosis is < 1 for the interaction between age, ε4 and each of the proteins above. (B) Interactions between ε4 and AA to DHA ratios in PC, PE, CE, and EA were associated with a significantly higher OR for the diagnosis of mTBI. As with protein biomarkers, interactions between age, ε4 and AA to DHA ratios in these lipid classes resulted in OR < 1 for a diagnosis of mTBI.

Fig. 3

The ratios of AA to DHA in lipid classes, when combined with the years since 1^st mTBI among ε4 carriers, are associated with differences in the EC thickness and volume. (A-E) Heatmaps showing beta-coefficients from linear regression modeling examining associations of the brain EC thickness and volume with different AA to DHA ratios and their interactions with years since 1^st mTBI and ε4. (A) Increasing AA to DHA ratios within the EA lipid class with increasing years since 1^st mTBI among ε4 carriers are positively associated with the GMV within the left EC. (B-D) There were no associations between AA to DHA ratios within DG, CE and PE and the brain EC thickness and volume. (E) Increasing AA to DHA ratios in PC with increasing years since 1^st mTBI among ε4 carriers are associated with the GMV in the left EC. *p < 0.05.

Fig. 4

Components 2 and 3 of PCA associated with r-mTBI predicted EC thickness and volume. Logistic regression analyses evaluated the individual and combined influence of PCA Component 2 with mTBI diagnosis and the presence of the ε4 allele on the odds of having lowered thickness and the volume of the EC. (A) Among ε4 carriers, PCA component 2, consisting of primarily AA and DHA-containing species in PC and PE along with DHA-containing species from CE and EA and AA-containing DG species, were associated with increased odds of having low GMV within the right EC, but the lower odds of GWV among those with ε4 and r-mTBI. (B) Among ε4 carriers with r-mTBI, PCA Component 3 containing AA and DHA species within PC and PE were associated with increased odds of having low WMV within the left EC. None of the PCAs were associated with the thickness of either the left or the right EC. *p < 0.05.

Fig. 5

A summary illustration of the hypothesis for mTBI and APOE-mediated AD risk and onset. The left panel of the image shows that mTBI, APOE and AA to DHA balance can independently present as AD risk. The right panel shows how these factors interact to precipitate the onset of AD.