Apolipoprotein E glycosylation and its role in Alzheimer's disease pathogenesis

Abstract Despite the critical importance of the O-glycoprotein apolipoprotein E (APOE) on Alzheimer’s disease (AD) risk, with the APOE4 isoform increasing risk compared to APOE3 and APOE2 reducing it, the precise mechanism behind this remains elusive. We have characterized APOE O-glycosylation using our quantitative glycoproteomic

targeted mass spectrometric approach showing that CSF and plasma APOE glycosylation differ greatly, particularly in the lipid binding domain. We have also developed a range of APOE binding assays to determine the impact of glycosylation on function. Our data using APOE from induced pluripotent stem cell (iPSC) derived

cells shows that glycosylation alters APOE binding properties. We also know that transferases, the enzymes that add the monosaccharides to glycans, wane with age and sialyltransferases are reduced in AD and there are differences in APOE modifications between APOE3.3 and APOE4.4 human brains. Together, this makes it

critical to fully understand APOE glycosylation. Thus, we hypothesize that APOE shows isoform-dependent glycosylation and that aberrant glycosylation alters APOE binding properties, exacerbating AD pathogenesis. We will use a range of glycobiology and iPSC techniques to address four Aims. Aim 1 will use

normal APOE isogenic iPSC-derived astrocytes and hepatocytes, the main producers of APOE. We will characterize their APOE glycoprofiles and associated binding properties, to gain an understanding of tissue and isoform-specific APOE glycosylation differences and their functional impacts. Transferase expression will be also

analyzed to further determine the mechanisms behind glycosylation differences. Aim 2 will compare astrocytes derived from healthy and AD APOE isogenic iPSCs and determine how AD alters APOE glycosylation; how this is altered between APOE isoforms, and how such changes impact APOE functions involved in AD pathogenesis.

Lipoprotein, receptor and Aβ binding will be compared. Finally, we will compare the APOE glycoprofiles of AD and normal human brain samples to our iPSC data. Aim 3 will use normal APOE isogenic iPSCs to model aging and AD by two methods to determine which more closely resembles the APOE glycosylation of AD. First by

reducing the specific sialyltransferase expression seen in aging and second by disrupting the Aβ environment by introducing a known APP mutation. This will elucidate how pathogenic glycosylation changes begin. Aim 4 will address if astrocytes with aberrant APOE glycosylation alter neuronal network activity and amyloid

accumulation by co-culture with iPSC-derived neurons. Our micro-electrode array (MEA) analyses have shown that the APOE genotype of astrocytes affects neuronal networks. We will use MEAs and measures of Aβ accumulation to determine the effect of these aberrantly glycosylated cell lines on neurons. Ultimately we will

have characterized normal and AD isoform-specific APOE glycosylation and defined its impact on APOE functions, especially those relating to AD pathogenesis. We will determine how aberrant APOE glycosylation is a promoter of Aβ accumulation, AD pathogenesis and neuron network degeneration.