DHA brain uptake and APOE4 status: a PET study with [1-¹¹C]-DHA

Apolipoprotein E ɛ4 (APOE4) genotype is the strongest genetic risk factor for late-onset or sporadic Alzheimer’s disease (AD). APOE proteins, the product of the APOE gene, have isoform-specific functions. For example, APOE’s affinity for the low-density lipoprotein receptor is known to differ between isoforms (APOE4 > APOE3 > APOE2) [1]. These differences have implications for the metabolism of APOE lipoprotein particles and the amount of lipid carried by APOE. In the brain, APOE forms high-density lipoprotein particles and participates in exchange of lipids between glial cells and neurons [2]. Clinical and animal studies indicate that brain APOE particle size and number differ by APOE genotype [3–5]. In plasma, APOE4 is catabolized faster with a plasma residence time of approximately half that of APOE3 [6].

Among brain lipids, the ω-3 polyunsaturated fatty acid (PUFA) docosahexaenoic acid (DHA, 22:6ω-3) may be of particular importance in AD pathogenesis. DHA forms up to 40% of fatty acids in certain gray matter lipids and is concentrated at synapses, where it plays a role in synaptic plasticity [7]. In embryonic neuronal cultures, DHA supplementation promotes neurite growth and synaptic protein expression [8]. Severe long-term dietary deficiency of DHA leads to learning impairment in animal models [9]. The brain also requires DHA for maintenance of neuronal membranes, production and clearance of β-amyloid 42, modulation of inflammation [10, 11], and cerebrovascular health [12]. We previously reported a direct association between lower serum DHA levels and greater cerebral amyloidosis in cognitively healthy older adults [13]. The lowest quartile of serum DHA was associated with significantly greater cerebral amyloid deposition, smaller entorhinal and hippocampal volumes, and worse nonverbal memory scores [13].

DHA’s incorporation into the brain can be assessed by positron emission tomography (PET) following intravenous infusion of carbon-11 ([1-¹¹C])-DHA using the incorporation coefficient k* [14]. k* represents multiple steps, including DHA diffusion from plasma to brain cells, intracellular DHA acylation to DHA-CoA by an acyl-coenzyme A (acyl-CoA) synthetase, and DHA transfer from DHA-CoA to membrane lysophospholipids by an acyltransferase [15]. k* is independent of changes in regional cerebral blood flow (rCBF). For example, rCBF can be doubled using CO₂ inhalation without changing k* [16]. The net rate of DHA incorporation from plasma (J _in) is the product of unesterified plasma DHA times k*. At steady state, J _in is equivalent to the net loss of DHA from the brain (J _out). Chronic dietary ω-3 PUFA deprivation leads to increased k* in the face of a 40-fold reduction in the rate of DHA loss (J _out) from the brain [17].

APOE genotype may influence the metabolism of DHA in the brain or its delivery to the brain, although brain DHA delivery may not directly depend on peripheral lipoproteins [18]. In humans, whole body DHA half-life was lower in APOE4 carriers than in noncarriers, which was attributed to greater liver oxidation of DHA [19]. Brain DHA levels were lower in older but not younger APOE4 targeted replacement (TR) mice than in age-matched APOE2 TR mice [20]. We found lower cerebrospinal fluid (CSF) DHA levels in older APOE4 carriers with mild AD after 18 months of DHA supplementation than in APOE4 noncarriers [21]. The goal of the present study was to explore the effect of APOE4 on [1-¹¹C]-DHA brain kinetics in a group of 22 healthy adults using PET.

In this exploratory post hoc analysis, we identified a significantly greater mean global gray matter DHA incorporation coefficient (k*) in APOE4 carriers compared with noncarriers. This difference was present in several brain regions, including the posterior cingulate cortex and the medial temporal lobe. Within the medial temporal lobe, higher DHA k* was prominent in the entorhinal cortex area. The simplest explanation for the significantly higher values of k* in APOE4 carriers is an increased incorporation by the brain from circulating unesterified DHA, replacing DHA that is either metabolized to bioactive products or lost to degradation. Given the small sample size and the exploratory nature of this study, these results are proof of concept and require additional validation.

Vandal et al. reported reduced brain DHA levels in older but not younger APOE4 mice compared with age-matched APOE2 TR mice [20]. We recently reported lower CSF DHA levels in older APOE4 carriers with AD after 18 months of DHA supplementation than in APOE4 noncarriers [21]. It is possible that the increased k* represents a compensatory mechanism in younger APOE4 carriers to cope with increased brain DHA loss and to maintain brain DHA levels. This mechanism might become impaired with aging, predisposing older APOE4 carriers to reduced brain DHA levels and increasing the risk for cognitive decline.

It is not possible on the basis of PET images to distinguish the exact metabolite explaining the higher incorporation of DHA in the brain. The equation for calculating k* assumes that all [1-¹¹C]-DHA is irreversibly trapped in the brain and that no radioactive metabolite other than [¹¹C]-CO₂ crosses the blood-brain barrier (BBB). This could result from more efficient transport of unesterified DHA across the BBB, increased activation of DHA to DHA-CoA by an acyl-CoA synthetase, greater esterification into brain membrane lipid by an acyltransferase, or decreased hydrolysis by phospholipase A₂ (PLA₂) [26]. Any one of these steps could be influenced by APOE genotype.

Several factors can alter k*. For example, k* was decreased in mice genetically lacking calcium-independent PLA₂β VIA [27], but it was increased when plasma and brain DHA concentrations were reduced by chronic dietary ω-3 PUFA deprivation in rats [17] or in subjects with chronic alcoholism during acute withdrawal of alcohol [22]. Moreover, the DHA transport coefficient was decreased with long-term high-DHA dietary consumption [28]. Therefore, differences in habitual intake of DHA may indirectly affect k*. To reduce variation in DHA intake in the present study, participants were instructed to avoid foods high in ω-3 PUFAs (e.g., seafood) 3 days preceding the PET scan, and they were limited to one caffeinated beverage per day. Beginning 24 h before the PET scan, they consumed standardized meals; in addition, they did not eat for 12 h prior to the scan. The differences in plasma DHA levels or DHA dietary intake were not significant by group.

The lower value of k* in the medial temporal cortex is consistent with previous reports for [1-¹¹C]- arachidonic acid and with values for rCBF [29, 30]. The data likely reflect the unique architecture of this region, although there is some effect of the PVE correction [24]. We previously reported that lower plasma levels of DHA were significantly associated with lower entorhinal brain volumes in older cognitively healthy adults with increased brain amyloidosis [13]. Higher ω-3 content of red blood cells was also associated with a lower rate of hippocampal atrophy [31]. Atrophy of this brain region predicts progression to AD [32]. Therefore, understanding the mechanisms that influence DHA metabolism in the medial temporal cortex is of particular relevance to AD.

Higher regional k* among this relatively young adult population of APOE4 carriers may provide one mechanism for increased regional brain activation observed in young adult APOE4 carriers [33–35]. One report demonstrated differences in myelin structure and gray matter volume in infants carrying the APOE4 allele [36]. Although APOE4 is associated with increased risk for memory decline and AD in older adults, several (but not all) studies suggest a behavioral advantage in APOE4 for younger carriers [37]. For example, in some studies, APOE4 has been associated with higher IQ scores [38] and a higher education level [39]. Advantageous effects of the APOE4 allele have also been found for memory-related functions in young animals. Hippocampal long-term potentiation (LTP) was enhanced at a young age in APOE4 TR mice compared with APOE4 noncarrier TR mice [40]. This LTP enhancement was age-dependent and disappeared in the adult mice. Mondadori et al. found an association of APOE4 with better episodic memory compared with APOE2 and APOE3 in 340 young, healthy persons [41]. Dennis et al. found enhanced functional connectivity of the medial temporal lobe with the posterior cingulate cortex in young adult APOE4 carriers [34]. Rusted et al. reported that the APOE4 in young adults was associated with improved attention and enhanced connectivity [35]. Filippini et al. reported increased default mode network coactivation in APOE4 carriers relative to noncarriers using resting-state functional MRI [33]. Combined, these findings suggest a state of increased brain activity decades prior to the onset of cognitive decline in APOE4 carriers. These reports support the “antagonistic pleiotropy” hypothesis in which cognitive advantages in younger adults support higher achievement and greater selection benefits, but may increase susceptibility to brain exhaustion and memory failure with age [42]. In this context, one possible interpretation of the higher k* is that APOE4 is associated with greater brain DHA loss and greater incorporation of DHA into the brain from plasma. These findings would suggest a beneficial response in cognitive function by increasing DHA consumption in APOE4 carriers in order to meet the greater metabolic demand for DHA in the brain. Researchers in several epidemiological studies and clinical trials have reported cognitive benefit from increasing DHA consumption in cognitively healthy APOE4 carriers [43]. This hypothesis merits additional investigation.

The study has several limitations. The sample size was small, and the study was a post hoc analysis of middle-aged, predominantly white adults. We also did not have sufficient participants to examine the effect of age or separate homozygous from heterozygous APOE4 carriers. DHA incorporation into the brain may not be dependent upon transport of peripheral lipoproteins; thus, our observed differences may be due to APOE4-related differences in transport across the BBB, intracellular transport, metabolism, or degradation processes. Unfortunately, data were not available to evaluate these hypotheses.

To our knowledge, this is the first study describing brain DHA incorporation coefficient in the context of the APOE4 allele and shows that brain regions implicated in the development of AD have different DHA incorporation coefficients, depending on APOE status. These findings support the development of novel DHA uptake imaging modalities such as [¹⁸F]-DHA to potentially accelerate the application of DHA imaging in clinical research. Knowledge of brain DHA metabolism will enhance understanding of how the APOE4 allele affects cognitive function and AD risk across the lifespan.