Long-term effects of maternal choline supplementation on CA1 pyramidal neuron gene expression in the Ts65Dn mouse model of Down syndrome and Alzheimer's disease

Abstract

Choline is critical for normative function of 3 major pathways in the brain, including acetylcholine biosynthesis, being a key mediator of epigenetic regulation, and serving as the primary substrate for the phosphatidylethanolamine N-methyltransferase pathway. Sufficient intake of dietary choline is critical for proper brain function and neurodevelopment. This is especially important for brain development during the perinatal period. Current dietary recommendations for choline intake were undertaken without critical evaluation of maternal choline levels. As such, recommended levels may be insufficient for both mother and fetus. Herein, we examined the impact of perinatal maternal choline supplementation (MCS) in a mouse model of Down syndrome and Alzheimer's disease, the Ts65Dn mouse relative to normal disomic littermates, to examine the effects on gene expression within adult offspring at ∼6 and 11 mo of age. We found MCS produces significant changes in offspring gene expression levels that supersede age-related and genotypic gene expression changes. Alterations due to MCS impact every gene ontology category queried, including GABAergic neurotransmission, the endosomal-lysosomal pathway and autophagy, and neurotrophins, highlighting the importance of proper choline intake during the perinatal period, especially when the fetus is known to have a neurodevelopmental disorder such as trisomy.-Alldred, M. J., Chao, H. M., Lee, S. H., Beilin, J., Powers, B. E., Petkova, E., Strupp, B. J., Ginsberg, S. D. Long-term effects of maternal choline supplementation on CA1 pyramidal neuron gene expression in the Ts65Dn mouse model of Down syndrome and Alzheimer's disease.

Keywords: early choline delivery; hippocampus; laser capture microdissection; microarray; trisomic.

Figure 1

Venn diagrams of gene expression level changes within CA1 pyramidal neurons due to MCS in Ts mice and 2N littermates at ∼6 and 11 mo old. A) In ∼6-MO mice, 328 genes are differentially regulated because of MCS in 2N mice and 309 in Ts mice, with 220 genes overlapping between genotypes, of which 51.4% are down-regulated and 48.6% are up-regulated. Only 1 overlapping gene (0.04%) has an expression level change in the opposite direction (mismatch) between genotypes. B) Mice at ∼11 mo display larger numbers of MCS-responsive genes, with 464 differentially regulated in 2N mice and 459 in Ts mice, respectively, with 396 genes overlapping between genotypes, of which 35.4% are down-regulated and 63.4% are up-regulated. Only 5 overlapping genes (1.3%) were mismatched. Arrows indicate down-regulated and up-regulated gene percentages, respectively, with ≠ indicating the low gene percentages that are in the opposite direction.

Figure 2

Venn diagrams of gene expression level changes within CA1 pyramidal neurons due to MCS in Ts mice and 2N littermates by comparing ages. A) In 2N mice, 328 genes are differentially regulated by MCS in young, ∼6-mo-old mice, and 464 genes are differentially regulated by MCS in older ∼11-mo-old mice, with 231 overlapping between age points. However, 121 of these MCS-responsive gene changes occur in the same direction, whereas 110 of these MCS-responsive gene changes occur in the opposite directions during aging. B) In Ts mice, 309 genes are differentially regulated by MCS in young, ∼6-mo-old mice, and 459 genes are differentially regulated by MCS in older ∼11-mo-old mice, with 213 overlapping between age points. Similar to 2N mice, less than half (88) of these MCS-responsive gene changes occur in the same direction, whereas 125 of these MCS-responsive gene changes occur in the opposite directions during aging. C) Color-coded heatmaps illustrating relative expression changes in 2N (left panel) and Ts (right panel) mice that have convergent expression because of MCS and age (top portion of the panel) or divergent (mismatched) expression because of MCS and age (bottom portion of the panel). Arrows indicate down-regulated and up-regulated gene totals, respectively.

Figure 3

Illustration of gene expression level changes within CA1 pyramidal neurons. Genes that were changed by MCS in 2N ∼6-mo-old mice (purple; 328 genes) and 2N ∼11-mo-old mice (red; 464 genes), Ts ∼6-mo-old mice (green; 309 genes) and Ts ∼11-mo-old mice (yellow; 459 genes) are shown, with overlapping genes affected in multiple cohorts (age + MCS or genotype + MCS). Genes displaying MCS effects regardless of age or genotype are shown in the central overlap of 129 genes. Of these 129 genes, less than half (53) are altered by MCS in the same direction (See Supplemental Fig. S2).

Figure 4

Depiction of MCS-responsive genes within CA1 pyramidal neurons in relevant GOC pathways. A) Pie chart illustrating expression level changes driven by MCS independent of age and genotype by GOC. B) Color-coded heatmap of key MCS-responsive genes that display convergent gene expression changes. Of these 53 transcripts, 13 are significantly down-regulated by MCS, and 40 genes are significantly up-regulated by MCS independently of age or genotype.

Figure 5

Validation of select expression level changes observed by single population microarray analysis in an independent cohort of ∼6-mo-old Ts and 2N mice. A) NanoString nCounter analysis corroborated microarray findings for Hspa8 and Mapk1 for 2N MCS-treated offspring and synaptophysin (Syp) for both 2N and Ts MCS-treated offspring, with a trend for Ntf5 gene expression from subregional CA1 dissections but no significant alterations in Rab5, and Rab7, confirming microarray findings. B) qPCR analysis validated Crebbp, Elavl1, and Gnaz custom-designed microarray observations because they were up-regulated in Ts mice and attenuated in Ts⁺ mice. Arrows indicate significant changes that are in the opposite direction of the single population microarray findings. N.s., nonsignificant. *P < 0.05, **P < 0.01, ***P < 0.001.