Plasmalogen deficiency and the Alzheimer’s disease risk of apolipoprotein ɛ4

Hällqvist J, Taanman J, Göteson A, Heywood WE, Schott JM, Hardy J, Landén M, Zetterberg H, Mills K, and Ginsberg L. (2026) Plasmalogen deficiency and the Alzheimer’s disease risk of apolipoprotein ɛ4. Brain Communications

The apolipoprotein E gene (APOE) encodes the apoE protein and is responsible for cholesterol transport and metabolism in the central nervous system and peripherally. APOE is the greatest genetic risk factor for late-onset Alzheimer’s disease (AD) and this risk is determined by the type and number of the alleles present. APOE ɛ2 is considered protective against the disease, APOE ɛ4 has the greatest risk, and APOE ɛ3 is considered the most common allele and is the baseline the other two are compared against. The number of each allele present is also important with risk level being dependent on the amount of ɛ2 and ɛ4 being present. Here is an example of the risk comparison based on allele presence: ɛ2ɛ2 < ɛ2ɛ3 < ɛ3ɛ3 < ɛ3ɛ4 < ɛ4ɛ4. The mechanism behind this risk is not clear and because of this Hällovist et al were interested in investigating the function of apoE in lipid biochemistry and whether these different isoforms that result from the alleles result in different phospholipid association properties and if any mediate risk of AD. The authors focused their lipidomic search on ethanolamine plasmalogens (PlsEtn) because a deficiency in PlsEtn has been identified in AD, and on phosphatidyl ethanolamines to use as a control. PlsEtn are a class of lipid that contain a vinyl-ether bond at the sn1 position on the glycerol backbone which causes this lipid class to have a more compact conformation. PlsEtn are important in the structure, fluidity, and function of the cell membrane, have antioxidant properties, and are essential for vesicular formation and cell signaling.

Cerebral spinal fluid (CSF) samples were obtained from the St. Göran bipolar project which consisted of adult patients with bipolar spectrum disorder from outpatient clinics and healthy controls. CSF samples from people with normal cognitive function were chosen to ensure that there would be no effect of neurodegeneration on the levels of lipids and proteins, and the groups had no significant difference in age, sex, or health status. CSF samples stratified by APOE alleles were randomly selected from this study (n = 5 ɛ3ɛ3, n = 4 ɛ3ɛ4, n = 5 ɛ4ɛ4). ApoE was purified from the CSF samples using immunoprecipitation (IP) then lipidomic and proteomic analyses of the precipitate were performed using mass spectrometry to determine what was bound to apoE. They monitored 40 species of PlsEtns and of those, 20 were associated with apoE-IP although C18:0/22:6 was most the dominant species and made up 40.3 mole % of total PlsEtn. There were 28 PtdEtn species monitored and 18 were associated with apoE-IP and C38:4 accounted for 26 mole % of total PtdEtn.

To compare the levels of PlsEtn and PtdEtn between the genotypes, PlsEtn/apoE and PtdEtn/apoE molar ratios were evaluated. A significant decrease in PlsEtn/apoE molar ratio was seen in 10/20 molecular species when comparing ɛ4ɛ4 to ɛ3ɛ3, however when looking at PtdEtn/apoE this trend was not observed. In the case of PtdEtn/apoE, three species were found to be decreased while six were increased in ɛ4ɛ4 compared to ɛ3ɛ3. In total, a 29.5% reduction in PlsEtn/apoE molar ratio was seen in the apoɛ4ɛ4 group compared to apoɛ3ɛ3 (p = 0.007). A biological gradient was able to be seen when PlsEtn/apoɛ4ɛ4, PlsEtn/apoɛ3ɛ4, and PlsEtn/apoɛ3ɛ3 were compared and a negative correlation was found between lipid/apoE molar ratio and amount of ɛ4 alleles present (0, 1, or 2 in ɛ3ɛ3, ɛ3ɛ4, and ɛ4ɛ4, respectively).

Hällqvist et al were interested in investigating the relationship between APOE allele and how its function affects lipid biochemistry, with particular interest in ethanolamine plasmalogens due to their association with Alzheimer’s disease. They used samples from cognitively healthy individuals to ensure that any changes seen were not due to neurodegeneration but instead associated with the alleles present. Since APOE ɛ4 is associated with the greatest risk while APOE ɛ3 is the most common allele, they chose ɛ3ɛ3, ɛ3ɛ4, and ɛ4ɛ4 as the genotypes to study. A biological gradient was detected when evaluating the PlsEtn/apoE molar ratios where PlsEtn/apoɛ3ɛ3 was found to have the greatest molar ratio and PlsEtn/apoɛ4ɛ4 the lowest in 10 out of 20 species, however when looking at PtdEtn/apoE molar ratios, this trend was not found, indicating that there is lipid specificity in the function of the different apoE isoforms since they are found to influence PlsEtn but not PtdEtn levels. One theory they had about this was that a possible mechanism for the increased risk of AD due to ɛ4ɛ4 is that this isoform is less able to restore PlsEtn deficiency in the brain that is known to occur in AD. Interestingly, when they compared PlsEtn/PtdEtn molar ratios in apoɛ4ɛ4 compared to apoɛ3ɛ3, the change detected is a similar extent to the difference between a brain with AD and a control brain. Further work looking into the protective effects of ɛ2 and any relationship with PlsEtn levels, in particular whether they remain the same as what was seen in apoɛ3ɛ3 or if they are greater than these levels, would be very interesting. Additional work that allows us to better understand the mechanisms behind each isoform and how they influence PlsEtn levels could be helpful in understanding Alzheimer’s disease and other potential treatments.