Maternal choline supplementation programs greater activity of the phosphatidylethanolamine N-methyltransferase (PEMT) pathway in adult Ts65Dn trisomic mice

Abstract

Maternal choline supplementation (MCS) induces lifelong cognitive benefits in the Ts65Dn mouse, a trisomic mouse model of Down syndrome and Alzheimer's disease. To gain insight into the mechanisms underlying these beneficial effects, we conducted a study to test the hypothesis that MCS alters choline metabolism in adult Ts65Dn offspring. Deuterium-labeled methyl-d9-choline was administered to adult Ts65Dn and disomic (2N) female littermates born to choline-unsupplemented or choline-supplemented Ts65Dn dams. Enrichment of d9-choline metabolites (derived from intact choline) and d3 + d6-choline metabolites [produced when choline-derived methyl groups are used by phosphatidylethanolamine N-methyltransferase (PEMT)] was measured in harvested tissues. Adult offspring (both Ts65Dn and 2N) of choline-supplemented (vs. choline-unsupplemented) dams exhibited 60% greater (P≤0.007) activity of hepatic PEMT, which functions in de novo choline synthesis and produces phosphatidylcholine (PC) enriched in docosahexaenoic acid. Higher (P<0.001) enrichment of PEMT-derived d3 and d6 metabolites was detected in liver, plasma, and brain in both genotypes but to a greater extent in the Ts65Dn adult offspring. MCS also yielded higher (P<0.05) d9 metabolite enrichments in liver, plasma, and brain. These data demonstrate that MCS exerts lasting effects on offspring choline metabolism, including up-regulation of the hepatic PEMT pathway and enhanced provision of choline and PEMT-PC to the brain.—Yan, J., Ginsberg, S. D., Powers, B., Alldred, M. J., Saltzman, A., Strupp, B. J., Caudill, M. A. Maternal choline supplementation programs greater activity of the phosphatidylethanolamine N-methyltransferase (PEMT) pathway in adult Ts65Dn trisomic mice.

Down syndrome (DS) is the most common genetic cause of intellectual disability and is often accompanied by behavioral disorders and attention deficits (1). Affecting ∼1 of 690 live births in the United States (2), DS is caused by the triplication of human chromosome 21 (HSA21), which arises from nondisjunction during meiosis (3). In addition to intellectual disability, individuals with DS develop Alzheimer's disease (AD) pathology, including amyloid plaques and neurofibrillary tangles, usually by the fourth decade of life (4–7). Because no specific treatment is available for the intellectual disability or dementia in individuals with DS (8), prognosis is generally poor.

The Ts65Dn mouse model has been developed to replicate the human DS condition via triplication of the distal portion of murine chromosome 16 (MMU16) orthologous to HSA21 (9). Ts65Dn mice manifest many of the biochemical, neuronal, and cognitive abnormalities observed in humans with DS (10–14). Specifically, these animals exhibit degeneration of the basal forebrain cholinergic neurons (BFCNs), with significant age-related reductions in total BFCN numbers and a reduction in choline acetyltransferase activity in the neocortex and hippocampus (11, 15–20), suggesting that some of the cognitive deficits observed in DS arise from impairments in cholinergic functioning.

Using the Ts65Dn mouse model of DS and AD, Strupp and colleagues (21, 22) have shown that supplementing the mother's diet with additional choline during pregnancy and lactation substantially lessens the neurocognitive dysfunction seen in this genetic disorder. Adult trisomic offspring born to mothers who consumed extra choline during the perinatal period performed significantly better on tasks assessing attention, spatial cognition, or emotional regulation compared to unsupplemented Ts65Dn mice and on some tests performed similarly to normal disomic (2N) littermate controls (21, 22). Comparable cognitive benefits have been reported in normal animals (23–27) and in animal models of seizure disorders (28, 29) and prenatal alcohol exposure (30, 31).

Although much remains to be learned about the mechanisms by which maternal choline supplementation (MCS) produces these beneficial effects in this DS mouse model, it is very likely that the basic mechanisms are the same as those that provide cognitive benefit in normal animals and other disease models. In normal rodents, pre- and/or early postnatal MCS has been shown to produce lasting changes in the cholinergic signaling system, hippocampal neurogenesis, and hippocampal long-term potentiation, all of which may contribute to the observed improvements in spatial memory and attention seen in these animals (reviewed in ref. 32). Similarly, studies with the Ts65Dn mouse model of DS have shown that MCS significantly increases hippocampal neurogenesis in the trisomic offspring and offers protection against the age-related degeneration of cholinergic neurons in the medial septal nucleus (22). Moreover, hippocampal neurogenesis and medial septal cholinergic neuron density each correlated significantly with the spatial cognition abilities of the offspring, suggesting functional relationships (21, 22). In addition, the persistent effects of MCS on cognition in adult offspring imply that an epigenetic mechanism mediated by the role of choline as a methyl donor may be involved (32).

To gain further insight into the mechanisms underlying MCS-induced benefits, the present study sought to test the hypotheses that choline metabolism differs between Ts65Dn mice and their 2N littermates and that MCS exerts lasting effects on choline metabolism in the adult Ts65Dn and 2N offspring. Isotopically labeled choline (i.e., methyl-d9-choline) was administered as a tracer in the drinking water of adult (16 mo) female trisomic (Ts65Dn) offspring and their 2N littermates born to dams who consumed the control or choline supplemented diet. As shown in Fig. 1, methyl-d9-choline is labeled with deuterium on all 3 methyl groups, which enables the tracing of the intact choline molecule (d9 metabolites) as well as the tracing of its methyl groups (d3 and d6 metabolites) (33, 34) in harvested tissues including liver, plasma, and discrete brain regions relevant to the neuropathology of DS and AD.

Figure 1.

Schematic of the metabolism of the orally consumed d9-choline tracer. The intact d9-choline tracer can be used to produce the d9-choline metabolites: d9-acetylcholine (d9-ACho), d9-betaine (d9-Bet), d9-phosphocholine (d9-PCho), d9-glycerophosphocholine (d9-GPC), and d9-phosphatidylcholine (d9-PC). Alternatively, the d9-choline tracer can be oxidized to d9-betaine, and one of its 3 methyl groups can be donated to homocysteine (Hcy), forming d3-methionine (d3-Met) and subsequently d3-S-adenosylmethionine (d3-SAM), the universal methyl donor. A main consumer of SAM is the phosphatidylethanolamine N-methyltransferase (PEMT) pathway, which mediates the sequential methylation of phosphatidylethanolamine (PE) to PC. Under the current labeling strategy, the PEMT pathway generated d3-PC and d6-PC. These PEMT-PC labeled metabolites can then undergo hydrolysis to release free choline (e.g., d3-Cho), which can be used to synthesize other metabolites (e.g., d3-ACho, d3-Bet, d3-PCho, and d3-GPC). Theoretically, d9-PC could also be generated by the PEMT pathway. However, the estimated amount of PEMT-derived d9-PC was negligible in the current study (as indicated by the very low enrichment of PEMT-derived d6-PC). Therefore, d9-PC production via the PEMT pathway is not depicted in the figure.

MATERIALS AND METHODS

Mice and diets

Breeder pairs (Ts65Dn female and C57Bl/6J Eicher × C3H/HeSnJ F1 male) were purchased from Jackson Laboratories (Bar Harbor, ME, USA) and mated at Cornell University (Ithaca, NY, USA). All animals were housed in microisolator cages (Ancare, Bellmore, NY, USA) in a temperature-controlled room (22–25°C and 70% humidity) with a 12-h light-dark cycle. At the time of mating, breeder pairs were randomized to either a diet containing standard choline levels (1.1 g choline chloride/kg diet; Dyets no.110098; Dyets, Bethlehem, PA, USA) or the same diet containing 5.0 g choline chloride/kg diet, which has been shown to produce lasting cognitive benefits in the offspring (22). These two levels of maternal choline intake continued until the pups were weaned at postnatal d 21. All pups were subsequently fed a standard chow diet and genotyped for the presence of the extra chromosome by Jackson Laboratories (35, 36).

These breeding pairs produce litters containing both trisomic (Ts65Dn) and disomic (2N) offspring. The present study used only female offspring because the male littermates were used for behavioral and cognitive testing (data not published). The study cohort included 8 disomic control (2N-C) and 7 trisomic control (Ts65Dn-C) offspring of 15 dams fed a control diet with standard choline levels, as well as 25 disomic MCS (2N-MCS) and 8 trisomic MCS (Ts65Dn-MCS) offspring of 15 choline-supplemented dams. For the 2N-C, Ts65Dn-C, and Ts65Dn-MCS groups (7–8 subjects/group), subjects were from different litters; for the 2N-MCS group with 25 subjects, ≤2 subjects were from the same breeder pairs.

At 16 mo of age, the animals were shipped to the Nathan Kline Institute for the tracer experiments. Methyl-d9-choline (1.6 mM; Cambridge Isotopes Laboratory, Andover, MA, USA) was provided as a tracer in the drinking water and was administered for 8 wk to the entire cohort as described previously (37). On completion of the tracer dosing, mice (18 mo of age) were overdosed with ketamine (13 mg/kg)/xylazine (83 mg/kg), blood was collected by cardiac puncture, mice were perfused with 1× PO₄ buffer (38), and tissues were collected and processed. In brief, blood was collected into vacutainer tubes containing 3.6 mg K₂-EDTA (Becton Dickinson, Franklin Lakes, NJ, USA), placed immediately on ice, processed for plasma within 4 h, and stored at −80°C (37). Liver was removed, cut into pieces, immediately frozen on dry ice, and stored at −80°C. Brain was harvested and dissected according to established protocols (39, 40) to extract basal forebrain, hippocampus, neocortex, cerebellum, and the rest of the brain, frozen on dry ice, and stored at −80°C. The samples (liver, plasma, and brain regions) were subsequently shipped (on dry ice) to the M.A.C. laboratory at Cornell University and stored at −80°C until analyzed for choline metabolite enrichments.

All protocols for the animal procedures were approved by the Institutional Animal Care and Use Committees of Cornell University, the Nathan Kline Institute, and the New York University Langone Medical Center and were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Measurement of plasma and tissue metabolite enrichment

Choline metabolites were extracted from the biological samples and quantified by liquid chromatography-tandem mass spectrometry (LC-MS/MS) according to Koc et al. (41) with modifications based on our instrumentation. The aqueous extract contained the water-soluble choline metabolites [acetylcholine, choline, betaine, glycerophosphocholine (GPC), and phosphocholine (PCho)], while the organic extract contained the lipid-soluble choline metabolites [phosphatidylcholine (PC), lysophosphatidylcholine (LPC), and sphingomyelin (SM)]. The LC-MS/MS system consisted of an Accela pump with degasser (Thermo, San Jose, CA, USA), a refrigerated Accela autosampler (Thermo), and a TSQ Quantum mass spectrometer (Thermo), which was equipped with an electrospray ionization source and operated in positive ion mode. Aqueous extract (10 μl) was injected onto a Prevail silica column (150×2.1 mm, 5 μm; Grace, Deerfield, IL, USA) with matching guard column (7.5×2.1 mm, 5 μm) for separation of the water-soluble choline metabolites. Mobile phases included solvent A (acetonitrile/water/ethanol/1 M ammonium acetate/glacial acetic acid, 800/127/68/3/2, v/v) and solvent B (acetonitrile/water/ethanol/1 M ammonium acetate/glacial acetic acid, 500/500/85/27/18, v/v). A 30 min gradient was used to achieve metabolite separation: t = 0 min, 100% solvent A; t = 2 min, 100% solvent A; t = 11 min, 55% solvent A; t = 14 min, 55% solvent A; t = 18 min, 40% solvent A; t = 19 min, 0% solvent A; t = 21.5 min, 0% solvent A; t = 25.5 min, 100% solvent A; t = 30 min, 100% solvent A. The metabolites of interest were detected by selected reaction monitoring of the following m/z ratio transitions: d0-aceytlcholine, 146/87; d3-acetylcholine, 149/87; d6-acetylcholine, 152/87; d9-acetylcholine, 155/87; d0-choline, 104/60; d3-choline, 107/63; d6-choline, 110/66; d9-choline, 113/69; d0-betaine, 118/59; d3-betaine, 121/62; d6-betaine, 124/65; d9-betaine, 127/68; d0-GPC, 258/104; d3-GPC, 261/107; d6-GPC, 264/110; d9-GPC, 267/113; d0-PCho, 184/125; d3-PCho, 187/125; d6-PCho, 190/125; d9-PCho, 193/125. Organic extract (10 μl) was injected onto an Adsorbosphere XL SI column (150×2.1 mm, 5 μm; Grace) with matching guard column (7.5×4.6 mm, 5 μm) for separation of the lipid-soluble choline metabolites. A 21 min gradient of the same solvents A and B was used: t = 0 min, 95% solvent A; t = 3 min, 100% solvent A; t = 8 min, 70% solvent A; t = 14 min, 40% solvent A; t = 16 min, 0% solvent A; t = 18 min, 0% solvent A; t = 20 min, 95% solvent A; t = 21 min, 95% solvent A. Because insource fragmentation of PC, SM, and LPC produces PCho as a common ion (regardless of fatty acid composition), PC, LPC, and SM were identified by selected ion monitoring of the PCho fragment, with m/z 184 for unlabeled d0 metabolites, m/z 187 for d3 metabolites, m/z 190 for d6 metabolites, and m/z 193 for d9 metabolites. The monitoring of the PCho fragment and the use of normal phase chromatography on the Adsorbosphere XL SI column provided the greatest sensitivity for quantifying isotopic enrichments but precluded the quantification of the molecular species of PC, SM, and LPC. No internal standards were used because only relative intensities of unlabeled and labeled metabolites (e.g., d0-, d3-, d6-, and d9-PC) are needed for calculations of fractional enrichments. Instruments were operated using Xcalibur 2.0.7 SP1 (Thermo), and data were processed with the Xcalibur Quan Browser to obtain peak areas under the chromatography curves for labeled and unlabeled choline metabolites.

Isotopic enrichment percentages [labeled metabolite/(labeled + unlabeled metabolites) × 100%] of the choline metabolites were calculated in liver, plasma, and discrete brain regions by using the peak area under the chromatography curve of the labeled metabolite divided by the total area of all isotopomers of the metabolite (37). These enrichment percentages were used to calculate 2 composite parameters to represent the overall metabolic use of choline-associated methyl groups by the phosphatidylethanolamine N-methyltransferase (PEMT) pathway [d3 + d6 enrichment = (total enrichment of d3- and d6-choline metabolites)/(number of d3- and d6-choline metabolites)] and the overall metabolic use of intact choline [d9 enrichment = (total enrichment of d9-choline metabolites)/(number of d9-choline metabolites)] for each tissue type. Because of the very low enrichment of d6-PC (a PEMT derivative), all of the d9-PC was assumed to have been derived from the cytidine diphosphate-choline (CDP-choline) pathway. Hepatic S-adenosylmethionine (d3-SAM) enrichment and PEMT activity were also calculated according to the principles of mass isotopomer distribution analysis (42, 43) using these equations: d3-SAM enrichment = (enrichment of d6-PC)/(enrichment of d3-PC + d6-PC); PEMT activity = (enrichment of d3-PC + d6-PC)/(enrichment of d3-SAM). Because d3- and d6-PC can be produced by the CDP-choline pathway using d3- and d6-choline released from PEMT-derived d3- and d6-PC (i.e., recycling of PEMT-PC), our calculated PEMT activity may overestimate actual activity. Nonetheless, this overestimation would transpire across all experimental groups, thereby enabling comparisons between genotypes and between treatment groups.

Statistical analysis

Data were logarithmically transformed to achieve normality and equal variance of residuals for all dependent variables; transformed data were used in subsequent analyses. The effect of genotype on the dependent variables (i.e., individual metabolite enrichments and composite parameters) was examined in the adult offspring of the unsupplemented dams using 1-factor ANOVA with genotype as the independent variable. The effect of MCS on the dependent variables (i.e., individual metabolite enrichments and composite parameters) was examined in all adult offspring using 2-factor ANOVA with MCS, genotype, and the interaction term (MCS×genotype) as the independent variables. If a significant interaction was detected, data were stratified by genotype, and the effect of MCS was tested using 1-factor ANOVA within the 2N and Ts65Dn offspring separately. Because body weight is a main determinant of water intake (44) (and thus tracer consumption), body weight was considered in the initial statistical analyses but was not retained in the final models because it was not a significant predictor.

Statistical analyses were conducted using SPSS 20 (IBM Corp., Armonk, NY, USA), and the Benjamini-Hochberg procedure was employed for false discovery rate (FDR) control at a level of 0.05 for the number of comparisons in each tissue type. Data are presented as estimated marginal means with 95% confidence intervals. FDR-adjusted P values are reported. Differences were considered statistically significant at values of P ≤ 0.05, and 0.05 < P ≤ 0.10 was indicative of trends.

RESULTS

Body weight at the end of tracer administration

A main effect of genotype (P=0.004) was detected for body weight with 22% greater weight observed in 2N (33±1 g) vs. Ts65Dn (27±1 g) mice. No effect of MCS (P=0.17) was detected for body weight, nor was an interaction between MCS and genotype detected (P=0.25).

Effect of genotype on individual enrichments of d3-, d6-, or d9-choline metabolites in adult offspring of unsupplemented dams

Liver

Ts65Dn (vs. 2N) offspring of unsupplemented dams exhibited significantly lower (P≤0.016) hepatic enrichments of all the labeled choline metabolites, including d3-betaine (−42%), d3-choline (−42%), d3-GPC (−40%,), d6-GPC (−74%), d3-PCho (−46%), d3-PC (−38%), d6-PC (−52%), d9-betaine (−63%), d9-choline (−50%), d9-GPC (−60%), d9-PCho (−60%), d9-LPC (−59%), d9-PC (−64%), and d9-SM (−44%) (Table 1), suggesting that hepatic choline uptake is reduced in Ts65Dn adult offspring. No differences (P=0.16) in hepatic PEMT activity were detected between genotypes with activities of 0.55% (0.43–0.68) vs. 0.67% (0.55–0.82) of total PC in Ts65Dn vs. 2N adult offspring, respectively. In addition, the activity of the hepatic CDP-choline pathway appeared to be unaffected by genotype, because the enrichment level of both its product (d9-PC) and its precursors (d9-choline and d9-PCho) were similarly reduced in Ts65Dn (vs. 2N) offspring.

Table 1.

Effects of genotype on choline metabolite enrichment in liver, plasma, and brain regions of adult disomic control (2N-C) and trisomic control (Ts65Dn-C) offspring born to unsupplemented control dams

Metabolite	2N-C, n = 8	Ts65Dn-C, n = 7	P	Metabolite	2N-C, n = 8	Ts65Dn-C, n = 7	P
Hepatic d3- and d6-choline metabolite enrichment (%)				Hepatic d9-choline metabolite enrichment (%)
d3-Betaine	1.1 (0.98–1.23)	0.6 (0.56–0.73)	≤0.016	d9-Betaine	0.4 (0.32–0.41)	0.1 (0.12–0.16)	≤0.016
d3-Choline	1.6 (1.39–1.82)	0.9 (0.79–1.09)	≤0.016	d9-Choline	0.9 (0.81–1.06)	0.5 (0.40–0.55)	≤0.016
d3-GPC	1.1 (0.97–1.27)	0.7 (0.57–0.77)	≤0.016	d9-GPC	0.8 (0.65–0.88)	0.3 (0.25–0.36)	≤0.016
d6-GPC	0.007 (0.004–0.013)	0.002 (0.001–0.004)	0.006	d9-PCho	0.5 (0.34–0.64)	0.2 (0.13–0.27)	0.001
d3-PCho	0.8 (0.70–1.01)	0.5 (0.36–0.56)	0.002	d9-LPC	0.6 (0.47–0.70)	0.2 (0.19–0.30)	≤0.016
d3-PC	1.0 (0.90–1.13)	0.6 (0.55–0.71)	≤0.016	d9-PC	0.5 (0.43–0.58)	0.2 (0.15–0.22)	≤0.016
d6-PC	0.015 (0.011–0.021)	0.007 (0.005–0.011)	0.006	d9-SM	0.7 (0.57–0.90)	0.4 (0.31–0.52)	0.004
Plasma d3- and d6-choline metabolite enrichment (%)				Plasma d9-choline metabolite enrichment (%)
d3-Betaine	1.0 (0.88–1.20)	0.6 (0.51–0.71)	≤0.01	d9-Betaine	0.3 (0.23–0.51)	0.2 (0.10–0.24)	0.012
d3-Choline	1.6 (1.47–1.82)	1.0 (0.88–1.11)	≤0.01	d9-Choline	3.0 (2.58–3.38)	1.8 (1.59–2.13)	≤0.01
d3-GPC	1.2 (1.04–1.29)	0.7 (0.62–0.79)	≤0.01	d9-GPC	0.7 (0.63–0.82)	0.3 (0.25–0.33)	≤0.01
d6-GPC	ND	ND		d9-PCho	ND	ND
d3-PCho	ND	ND		d9-LPC	0.6 (0.54–0.70)	0.2 (0.20–0.26)	≤0.01
d3-PC	0.9 (0.83–1.06)	0.6 (0.50–0.65)	≤0.01	d9-PC	0.5 (0.47–0.60)	0.2 (0.15–0.20)	≤0.01
d6-PC	ND	ND		d9-SM	0.7 (0.59–0.72)	0.4 (0.35–0.43)	≤0.01
Basal forebrain d3- and d6-choline metabolite enrichment (%)				Basal forebrain d9-choline metabolite enrichment (%)
d3-ACho	1.3 (1.20–1.50)	1.2 (1.07–1.36)	0.6	d9-ACho	0.7 (0.61–0.83)	0.7 (0.63–0.88)	0.9
d3-Betaine	1.3 (1.20–1.37)	1.1 (1.05–1.21)	0.3	d9-Betaine	0.4 (0.35–0.44)	0.3 (0.30–0.38)	0.3
d3-Choline	2.5 (2.30–2.75)	2.3 (2.06–2.49)	0.5	d9-Choline	1.4 (1.26–1.59)	1.4 (1.24–1.60)	0.9
d3-GPC	1.5 (1.36–1.60)	1.4 (1.28–1.52)	0.7	d9-GPC	1.1 (0.99–1.25)	1.2 (1.05–1.36)	0.7
d6-GPC	0.025 (0.020–0.031)	0.024 (0.019–0.030)	0.9	d9-PCho	0.6 (0.54–0.73)	0.7 (0.56–0.77)	0.9
d3-PCho	0.7 (0.36–1.20)	0.6 (0.30–1.08)	0.9	d9-LPC	ND	ND
d3-PC	1.2 (1.09–1.33)	1.2 (1.07–1.32)	0.9	d9-PC	0.9 (0.82–1.04)	1.0 (0.92–1.19)	0.5
d6-PC	0.023 (0.020–0.027)	0.021 (0.018–0.025)	0.7	d9-SM	0.7 (0.63–0.84)	0.9 (0.80–1.09)	0.2
Hippocampus d3- and d6-choline metabolite enrichment (%)				Hippocampus d9-choline metabolite enrichment (%)
d3-ACho	1.2 (1.03–1.39)	1.1 (0.97–1.34)	0.9	d9-ACho	0.7 (0.57–0.84)	0.8 (0.63–0.96)	0.8
d3-Betaine	1.3 (1.12–1.40)	1.1 (0.97–1.23)	0.8	d9-Betaine	0.3 (0.28–0.40)	0.3 (0.26–0.39)	0.9
d3-Choline	2.3 (2.01–2.52)	2.0 (1.80–2.30)	0.9	d9-Choline	1.4 (1.24–1.61)	1.4 (1.19–1.57)	0.9
d3-GPC	1.4 (1.33–1.57)	1.4 (1.25–1.50)	0.8	d9-GPC	1.1 (0.98–1.26)	1.2 (1.08–1.41)	0.7
d6-GPC	0.020 (0.014–0.029)	0.023 (0.016–0.034)	0.9	d9-PCho	0.7 (0.61–0.81)	0.7 (0.61–0.83)	0.9
d3-PCho	1.1 (0.97–1.22)	1.0 (0.87–1.11)	0.8	d9-LPC	ND	ND
d3-PC	1.2 (1.07–1.30)	1.2 (1.05–1.29)	0.9	d9-PC	0.9 (0.79–1.01)	1.0 (0.91–1.17)	0.6
d6-PC	0.022 (0.018–0.027)	0.021 (0.017–0.026)	0.9	d9-SM	0.7 (0.63–0.81)	0.9 (0.78–1.02)	0.4
Neocortex d3- and d6-choline metabolite enrichment (%)				Neocortex d9-choline metabolite enrichment (%)
d3-ACho	0.8 (0.69–1.04)	0.7 (0.59–0.92)	0.6	d9-ACho	0.5 (0.35–0.64)	0.4 (0.29–0.55)	0.6
d3-Betaine	1.2 (1.05–1.27)	1.0 (0.88–1.08)	0.2	d9-Betaine	0.3 (0.26–0.36)	0.2 (0.20–0.28)	0.2
d3-Choline	2.2 (2.03–2.43)	2.0 (1.85–2.24)	0.4	d9-Choline	1.4 (1.26–1.61)	1.5 (1.33–1.73)	0.6
d3-GPC	1.5 (1.33–1.62)	1.4 (1.23–1.52)	0.6	d9-GPC	1.2 (1.03–1.30)	1.2 (1.09–1.40)	0.7
d6-GPC	0.021 (0.013–0.036)	0.011 (0.006–0.019)	0.3	d9-PCho	0.6 (0.46–0.86)	0.5 (0.32–0.63)	0.4
d3-PCho	1.1 (0.88–1.29)	0.7 (0.61–0.91)	0.3	d9-LPC	ND	ND
d3-PC	1.2 (1.13–1.35)	1.2 (1.10–1.33)	0.9	d9-PC	0.9 (0.81–1.04)	1.0 (0.91–1.19)	0.4
d6-PC	0.023 (0.019–0.028)	0.022 (0.018–0.027)	0.9	d9-SM	0.7 (0.63–0.82)	0.9 (0.77–1.03)	0.1
Cerebellum d3- and d6-choline metabolite enrichment (%)				Cerebellum d9-choline metabolite enrichment (%)
d3-ACho	0.8 (0.58–1.14)	0.6 (0.45–0.91)	0.5	d9-ACho	0.3 (0.20–0.51)	0.3 (0.16–0.42)	0.7
d3-Betaine	1.3 (1.19–1.39)	1.0 (0.96–1.14)	0.034	d9-Betaine	0.4 (0.32–0.42)	0.3 (0.26–0.35)	0.1
d3-Choline	2.3 (2.13–2.56)	2.0 (1.84–2.24)	0.1	d9-Choline	1.4 (1.27–1.61)	1.4 (1.22–1.58)	0.8
d3-GPC	1.5 (1.37–1.61)	1.3 (1.24–1.47)	0.2	d9-GPC	1.2 (1.03–1.30)	1.2 (1.03–1.33)	0.9
d6-GPC	0.027 (0.022–0.034)	0.020 (0.016–0.025)	0.1	d9-PCho	0.8 (0.68–0.89)	0.6 (0.54–0.73)	0.2
d3-PCho	1.2 (1.08–1.34)	0.9 (0.85–1.06)	0.043	d9-LPC	ND	ND
d3-PC	1.2 (1.10–1.32)	1.1 (1.03–1.24)	0.5	d9-PC	0.9 (0.80–1.01)	1.0 (0.86–1.12)	0.5
d6-PC	0.024 (0.020–0.028)	0.022 (0.018–0.026)	0.7	d9-SM	0.5 (0.45–0.57)	0.6 (0.55–0.70)	0.1
Rest of brain d3- and d6-choline metabolite enrichment (%)				Rest of brain d9-choline metabolite enrichment (%)
d3-ACho	1.4 (1.20–1.60)	1.3 (1.12–1.52)	0.9	d9-ACho	0.8 (0.67–0.92)	0.8 (0.64–0.91)	0.9
d3-Betaine	1.3 (1.21–1.43)	1.2 (1.06–1.27)	0.2	d9-Betaine	0.4 (0.32–0.45)	0.4 (0.30–0.43)	0.9
d3-Choline	2.3 (2.13–2.54)	2.2 (1.97–2.37)	0.8	d9-Choline	1.4 (1.22–1.52)	1.4 (1.25–1.59)	0.9
d3-GPC	1.4 (1.32–1.56)	1.3 (1.23–1.48)	0.9	d9-GPC	1.1 (0.95–1.22)	1.2 (1.01–1.31)	0.9
d6-GPC	0.026 (0.020–0.033)	0.027 (0.020–0.035)	0.9	d9-PCho	0.8 (0.71–0.89)	0.8 (0.68–0.87)	0.9
d3-PCho	1.2 (1.08–1.28)	1.0 (0.93–1.12)	0.3	d9-LPC	ND	ND
d3-PC	1.2 (1.12–1.32)	1.2 (1.10–1.31)	0.9	d9-PC	0.9 (0.83–1.05)	1.1 (1.05–1.22)	0.4
d6-PC	0.024 (0.020–0.029)	0.022 (0.018–0.027)	0.9	d9-SM	0.7 (0.60–0.77)	0.9 (0.78–1.03)	0.1

Data were analyzed with 1-factor ANOVA and are presented as estimated marginal means with 95% confidence intervals. ND, not detected.

Plasma

Comparable to the labeling pattern observed in the liver, Ts65Dn (vs. 2N) offspring of unsupplemented dams exhibited significantly lower (P≤0.012) plasma enrichment of all the labeled metabolites, including d3-betaine (−41%), d3-choline (−39%), d3-GPC (−39%), d3-PC (−40%), d9-betaine (−54%), d9-choline (−37%), d9-GPC (−60%), d9-LPC (−63%), d9-PC (−67%), and d9-SM (−40%) (Table 1). d3-PCho, d6-GPC, d6-PC, and d9-PCho were not detected in plasma in either genotype.

Brain

In contrast to the labeling pattern observed in liver and plasma, enrichment of the choline metabolites did not differ (P>0.1) between genotypes in the examined brain regions, with the exception of the cerebellum, where d3-betaine and d3-PCho were ∼20% lower (P≤0.043) in the Ts65Dn (vs. 2N) adult offspring of unsupplemented dams (Table 1).

Effect of genotype on the overall enrichment of d3- and d6- or d9-choline metabolites in adult offspring of unsupplemented dams

Ts65Dn (vs. 2N) offspring of unsupplemented dams exhibited significantly lower (P<0.001) d3 and d6 enrichment and significantly lower (P<0.001) d9 enrichment in liver (−41 and −56%, respectively) and plasma (−40 and −46%, respectively) (Fig. 2). In contrast, d3 and d6 enrichment and d9 enrichment did not differ (P≥0.07) between genotypes in the examined brain regions, with the exception of cerebellum d3 and d6 enrichment, which was 14% lower (P=0.025) in the Ts65Dn (vs. 2N) adult offspring of unsupplemented dams (Fig. 2).

Figure 2.

Effects of genotype on overall d3 and d6enrichment (A) and overall d9 enrichment (B) in liver, plasma, and brain regions of adult disomic control (2N-C) and trisomic control (Ts65Dn-C) offspring born to unsupplemented control dams. Data are estimated marginal means with 95% confidence intervals. *P ≤ 0.05, ***P ≤ 0.001.

Effect of MCS during the perinatal period on individual enrichment of d3-, d6-, or d9-choline metabolites in the adult offspring

Liver

A main effect of MCS on d9-choline metabolite enrichment (except d9-betaine) was detected (P≤0.007) with higher enrichments of d9-choline (+43%), d9-GPC (+58%), d9-PCho (+91%), d9-LPC (+48%), d9-PC (+58%), and d9-SM (+41%) in the adult offspring of supplemented (vs. unsupplemented) dams (Table 2). Hepatic PEMT activity was also 60% higher (P≤0.007, main effect) in the adult offspring of supplemented vs. unsupplemented dams. For the d3- and d6-choline metabolites, as well as d9-betaine, MCS interacted with genotype (P≤0.041) to affect hepatic enrichments (Table 3). Ts65Dn adult offspring of supplemented (vs. unsupplemented) dams exhibited a markedly significant increase (P≤0.009) in enrichment of d3-betaine (+140%), d3-choline (+161%), d3-GPC (+155%), d6-GPC (+640%), d3-PCho (+201%), d3-PC (+159%), d6-PC (+249%), and d9-betaine (+148%); whereas 2N adult offspring of supplemented (vs. unsupplemented) dams exhibited a less robust but still significantly higher (P≤0.018) enrichment in select metabolites, including d3-choline (+29%), d3-GPC (+30%), d3-PC (+30%), and d9-betaine (+42%).

Table 2.

Effects of MCS on choline metabolite enrichment in liver, plasma, and brain regions of adult disomic control (2N-C) and trisomic control (Ts65Dn-C) offspring born to unsupplemented control dams vs. 2N-MCS and Ts65Dn-MCS offspring born to MCS dams

Metabolite	2N-C + Ts65Dn-C, n = 15	2N-MCS + Ts65Dn-MCS, n = 33	P	Metatolite	2N-C + Ts65Dn-C, n = 15	2N-MCS + Ts65Dn-MCS, n = 33	P
Hepatic d9-choline metabolite enrichment (%)				Plasma d9-choline metabolite enrichment (%)
d9-Betaine	See Table 3	See Table 3		d9-Betaine	0.2 (0.17–0.33)	0.3 (0.25–0.43)	0.1
d9-Choline	0.7 (0.56–0.78)	0.9 (0.84–1.06)	0.001	d9-Choline	2.3 (2.04–2.68)	2.9 (2.60–3.21)	0.019
d9-GPC	0.5 (0.40–0.58)	0.8 (0.67–0.87)	≤0.007	d9-GPC	0.5 (0.37–0.56)	0.6 (0.54–0.73)	0.021
d9-PCho	0.3 (0.24–0.38)	0.6 (0.48–0.68)	≤0.007	d9-PCho	ND	ND
d9-LPC	0.4 (0.31–0.44)	0.6 (0.49–0.63)	0.001	d9-LPC	0.4 (0.31–0.45)	0.6 (0.48–0.64)	0.003
d9-PC	0.3 (0.26–0.37)	0.5 (0.43–0.56)	≤0.007	d9-PC	0.3 (0.26–0.37)	0.5 (0.40–0.53)	0.002
d9-SM	0.5 (0.44–0.65)	0.8 (0.65–0.87)	0.006	d9-SM	0.5 (0.44–0.58)	0.7 (0.64–0.80)	0.006
Basal forebrain d3- and d6-choline metabolite enrichment (%)				Basal forebrain d9-choline metabolite enrichment (%)
d3-ACho	1.3 (1.11–1.48)	1.6 (1.41–1.74)	0.049	d9-ACho	0.7 (0.61–0.87)	0.8 (0.69–0.89)	0.5
d3-Betaine	1.2 (1.09–1.33)	1.6 (1.50–1.74)	≤0.017	d9-Betaine	0.4 (0.32–0.42)	0.4 (0.40–0.49)	0.058
d3-Choline	2.4 (2.17–2.65)	3.1 (2.85–3.31)	≤0.017	d9-Choline	1.4 (1.23–1.60)	1.7 (1.53–1.86)	0.057
d3-GPC	1.4 (1.31–1.57)	1.9 (1.75–2.00)	≤0.017	d9-GPC	1.1 (1.00–1.32)	1.3 (1.22–1.49)	0.088
d6-GPC	0.025 (0.019–0.032)	0.028 (0.023–0.034)	0.5	d9-PCho	0.6 (0.54–0.76)	0.7 (0.65–0.84)	0.2
d3-PCho	0.6 (0.48–0.79)	1.3 (1.09–1.57)	≤0.017	d9-LPC	ND	ND
d3-PC	1.2 (1.09–1.32)	1.6 (1.45–1.67)	≤0.017	d9-PC	1.0 (0.86–1.12)	1.1 (1.04–1.26)	0.093
d6-PC	0.022 (0.019–0.026)	0.031 (0.028–0.035)	0.003	d9-SM	0.8 (0.72–0.94)	0.9 (0.84–1.02)	0.2
Hippocampus d3- and d6-choline metabolite enrichment (%)				Hippocampus d9-choline metabolite enrichment (%)
d3-ACho	1.2 (1.04–1.32)	1.6 (1.46–1.74)	≤0.017	d9-ACho	0.7 (0.62–0.87)	0.9 (0.77–0.99)	0.1
d3-Betaine	1.2 (1.06–1.30)	1.5 (1.43–1.66)	≤0.017	d9-Betaine	0.3 (0.28–0.38)	0.4 (0.33–0.42)	0.2
d3-Choline	2.2 (1.95–2.38)	2.9 (2.67–3.10)	≤0.017	d9-Choline	1.4 (1.21–1.59)	1.7 (1.53–1.87)	0.039
d3-GPC	1.4 (1.28–1.55)	1.9 (1.75–2.01)	≤0.017	d9-GPC	1.2 (1.02–1.33)	1.4 (1.25–1.52)	0.061
d6-GPC	0.022 (0.018–0.028)	0.031 (0.026–0.037)	0.03	d9-PCho	0.7 (0.60–0.82)	0.9 (0.79–1.00)	0.03
d3-PCho	1.0 (0.94–1.15)	1.4 (1.32–1.53)	≤0.017	d9-LPC	ND	ND
d3-PC	1.2 (1.07–1.29)	1.5 (1.40–1.61)	≤0.017	d9-PC	1.0 (0.84–1.09)	1.1 (1.01–1.23)	0.081
d6-PC	0.022 (0.018–0.025)	0.031 (0.028–0.035)	0.002	d9-SM	0.8 (0.70–0.91)	0.9 (0.79–0.97)	0.2
Neocortex d3- and d6-choline metabolite enrichment (%)				Neocortex d9-choline metabolite enrichment (%)
d3-ACho	0.8 (0.69–0.92)	1.2 (1.07–1.34)	≤0.015	d9-ACho	0.4 (0.29–0.63)	0.5 (0.40–0.72)	0.3
d3-Betaine	1.1 (0.97–1.18)	1.4 (1.34–1.55)	≤0.015	d9-Betaine	0.3 (0.23–0.32)	0.4 (0.35–0.43)	≤0.015
d3-Choline	2.1 (1.93–2.36)	2.9 (2.69–3.11)	≤0.015	d9-Choline	1.5 (1.28–1.68)	1.7 (1.57–1.93)	0.066
d3-GPC	1.4 (1.30–1.56)	1.9 (1.77–2.03)	≤0.015	d9-GPC	1.2 (1.04–1.36)	1.4 (1.27–1.55)	0.074
d6-GPC	See Table 3	See Table 3		d9-PCho	0.5 (0.43–0.67)	0.7 (0.63–0.87)	0.035
d3-PCho	See Table 3	See Table 3		d9-LPC	ND	ND
d3-PC	1.2 (1.11–1.34)	1.6 (1.47–1.69)	≤0.015	d9-PC	1.0 (0.85–1.11)	1.1 (1.04–1.26)	0.07
d6-PC	0.023 (0.019–0.027)	0.032 (0.025–0.034)	0.002	d9-SM	0.8 (0.70–0.91)	0.9 (0.83–1.00)	0.1
Cerebellum d3- and d6-choline metabolite enrichment (%)				Cerebellum d9-choline metabolite enrichment (%)
d3-ACho	0.7 (0.55–0.95)	1.0 (0.81–1.22)	0.089	d9-ACho	0.3 (0.22–0.38)	0.4 (0.35–0.54)	0.043
d3-Betaine	1.2 (1.05–1.30)	1.6 (1.46–1.72)	≤0.017	d9-Betaine	0.3 (0.29–0.39)	0.4 (0.36–0.45)	0.085
d3-Choline	2.2 (2.00–2.40)	2.9 (2.75–3.16)	≤0.017	d9-Choline	1.4 (1.23–1.61)	1.7 (1.55–1.90)	0.037
d3-GPC	1.4 (1.30–1.56)	1.9 (1.79–2.05)	≤0.017	d9-GPC	1.2 (1.01–1.34)	1.4 (1.29–1.59)	0.04
d6-GPC	0.023 (0.017–0.032)	0.031 (0.024–0.039)	0.2	d9-PCho	0.7 (0.57–0.86)	0.8 (0.71–0.97)	0.2
d3-PCho	1.1 (0.95–1.22)	1.5 (1.32–1.60)	≤0.017	d9-LPC	ND	ND
d3-PC	1.2 (1.06–1.29)	1.5 (1.43–1.65)	≤0.017	d9-PC	0.9 (0.81–1.08)	1.1 (0.48–0.65)	0.085
d6-PC	0.023 (0.019–0.027)	0.032 (0.029–0.036)	0.002	d9-SM	0.6 (0.48–0.65)	0.6 (0.55–0.68)	0.3
Rest of brain d3- and d6-choline metabolite enrichment (%)				Rest of brain d9-choline metabolite enrichment (%)
d3-ACho	1.3 (1.22–1.50)	2.0 (1.85–2.15)	≤0.017	d9-ACho	0.8 (0.66–0.89)	1.0 (0.90–1.12)	0.009
d3-Betaine	1.2 (1.13–1.36)	1.7 (1.61–1.85)	≤0.017	d9-Betaine	0.4 (0.32–0.42)	0.5 (0.44–0.54)	0.004
d3-Choline	2.2 (2.04–2.48)	3.0 (2.77–3.20)	≤0.017	d9-Choline	1.4 (1.21–1.57)	1.6 (1.47–1.79)	0.059
d3-GPC	1.4 (1.27–1.53)	1.8 (1.71–1.96)	≤0.017	d9-GPC	1.1 (0.97–1.27)	1.3 (1.19–1.45)	0.062
d6-GPC	0.026 (0.022–0.032)	0.039 (0.034–0.045)	0.002	d9-PCho	0.8 (0.67–0.89)	0.9 (0.84–1.04)	0.05
d3-PCho	1.1 (1.00–1.21)	1.5 (1.40–1.62)	≤0.017	d9-LPC	ND	ND
d3-PC	1.2 (1.10–1.32)	1.6 (1.47–1.67)	≤0.017	d9-PC	1.0 (0.87–1.14)	1.1 (1.04–1.26)	0.1
d6-PC	0.023 (0.020–0.027)	0.034 (0.030–0.038)	≤0.017	d9-SM	0.8 (0.67–0.89)	0.9 (0.79–0.96)	0.2

Data were analyzed with 2-factor ANOVA (MCS, genotype, MCS×genotype). MCS did not interact with genotype (P>0.13) to affect these dependent variables; thus, the main effect of MCS is presented. Data are estimated marginal means with 95% confidence intervals. ND, not detected.

Table 3.

Effects of the interaction of genotype and MCS on choline metabolite enrichments in liver, plasma, and brain regions of adult disomic control (2N-C) and trisomic control (Ts65Dn-C) offspring born to unsupplemented control dams vs. 2N-MCS and Ts65Dn-MCS offspring born to MCS dams

Metabolite	2N-C, n = 8	2N-MCS, n = 25	P	Ts65Dn-C, n = 7	Ts65Dn-MCS, n = 8	P
Hepatic d3- and d6-choline metabolite enrichment (%)
d3-Betaine	1.1 (0.94–1.29)	1.3 (1.16–1.39)	0.2	0.6 (0.53–0.77)	1.5 (1.30–1.81)	≤0.009
d3-Choline	1.6 (1.37–1.85)	2.0 (1.88–2.23)	0.011	0.9 (0.75–1.15)	2.4 (2.02–2.92)	≤0.009
d3-GPC	1.1 (0.97–1.28)	1.4 (1.33–1.56)	0.014	0.7 (0.53–0.84)	1.7 (1.39–2.07)	≤0.009
d6-GPC	0.007 (0.004–0.013)	0.009 (0.006–0.013)	0.6	0.002 (0.001–0.004)	0.014 (0.008–0.026)	≤0.009
d3-PCho	0.8 (0.71–0.99)	1.0 (0.94–1.14)	0.063	0.5 (0.35–0.59)	1.4 (1.09–1.71)	≤0.009
d3-PC	1.0 (0.87–1.16)	1.3 (1.21–1.42)	0.018	0.6 (0.51–0.77)	1.6 (1.35–1.94)	≤0.009
d6-PC	0.015 (0.012–0.020)	0.019 (0.016–0.021)	0.2	0.007 (0.005–0.011)	0.026 (0.018–0.038)	0.001
Hepatic d9-choline metabolite enrichment (%)
d9-Betaine	0.4 (0.30–0.44)	0.5 (0.46–0.58)	0.012
d3-Betaine				0.1 (0.10–0.19)	0.3 (0.25–0.45)	0.001
Plasma d3- and d6-choline metabolite enrichment (%)
d3-Betaine	1.0 (0.81–1.31)	1.1 (0.95–1.26)	0.7	0.6 (0.49–0.73)	1.5 (1.22–1.82)	≤0.004
d3-Choline	1.6 (1.35–1.98)	1.8 (1.57–1.97)	0.5	1.0 (0.81–1.21)	2.4 (1.97–2.94)	≤0.004
d3-GPC	1.2 (0.99–1.35)	1.4 (1.24–1.50)	0.1	0.7 (0.57–0.86)	1.7 (1.36–2.04)	≤0.004
d6-GPC	ND	ND		ND	ND
d3-PCho	ND	ND		ND	ND
d3-PC	0.9 (0.81–1.09)	1.2 (1.08–1.29)	0.056	0.6 (0.46–0.70)	1.4 (1.17–1.78)	≤0.004
d6-PC	ND	ND		ND	ND
Neocortex d3- and d6-choline metabolite enrichments (%)
d6-GPC	0.021 (0.015–0.031)	0.025 (0.020–0.030)	0.5	0.011 (0.006–0.020)	0.040 (0.023–0.068)	0.004
d3-PCho	1.1 (0.90–1.26)	1.3 (1.14–1.39)	0.2	0.7 (0.59–0.94)	1.5 (1.17–1.82)	0.002

Data were analyzed with 2-factor ANOVA (MCS, genotype, MCS×genotype). Significant (P≤0.041) and borderline significant (P≤0.068) MCS and genotype interactions were detected for these dependent variables; thus, data were stratified by genotype, and the effect of MCS is presented separately for each genotype (2N-C vs. 2N-MCS; Ts65Dn-C vs. Ts65Dn-MCS). Data are estimated marginal means with 95% confidence intervals. ND, not detected.

Because the Ts65Dn genotype perturbed hepatic choline metabolism, we examined whether these abnormalities were normalized by MCS. Compared to the 2N adult offspring of unsupplemented dams, Ts65Dn adult offspring of supplemented dams exhibited similar (P≥0.2) or higher (P≤0.043) choline metabolite enrichment, as well as significantly higher PEMT activity (+57%, P=0.016) (Supplemental Table S1).

Plasma

The enrichment patterns of choline metabolites in plasma mirrored those observed in liver. A main effect of MCS on d9-choline metabolite enrichment (except d9-betaine) was detected (P≤0.021) with higher enrichment of d9-choline (+24%), d9-GPC (+38%), d9-LPC (+48%), d9-PC (+49%), and d9-SM (+43%) in the adult offspring of supplemented (vs. unsupplemented) dams (Table 2). For the d3-choline metabolites, MCS interacted with genotype (P≤0.01) to affect plasma enrichment (Table 3). Ts65Dn adult offspring of supplemented (vs. unsupplemented) dams exhibited substantially significantly higher (P≤0.004) enrichment of d3-betaine (+147%), d3-choline (+143%), d3-GPC (+138%), and d3-PC (+155%), whereas 2N adult offspring of supplemented (vs. unsupplemented) dams tended to exhibit higher (P=0.056) enrichment of d3-PC (+26%) only. Finally, MCS normalized choline metabolism in the trisomic offspring. Specifically, Ts65Dn adult offspring of supplemented dams exhibited similar (P≥0.068) or higher (P≤0.05) plasma choline metabolite enrichments than the unsupplemented 2N offspring, with the exception of d9-PC, which remained lower (−40%; P=0.048) than the unsupplemented 2N mice (Supplemental Table S1).

Brain

A main effect of MCS was detected (P<0.05) on the majority of d3-choline metabolites and, to a lesser extent, on the d9-choline metabolites. Specifically, higher choline metabolite enrichment was detected in 93% of the d3 metabolites (i.e., 28 of 30) and 29% of the d9 metabolites (i.e., 10 of 35) in the adult offspring of supplemented (vs. unsupplemented) dams (Table 2). Notably, while the enrichment of all the d3-choline metabolites was higher in the basal forebrain, none of the d9-choline metabolite enrichment levels achieved statistical significance (Table 2). MCS also tended to interact with genotype (P≤0.068) to influence d3-PCho and d6-GPC in the neocortex (Table 3). Specifically, Ts65Dn adult offspring of supplemented (vs. unsupplemented) dams exhibited substantially significantly higher (P≤0.004) enrichment of d6-GPC (+263%) and d3-PCho (+96%) whereas no differences (P>0.17) in the enrichments of these metabolites were detected in 2N adult offspring of supplemented (vs. unsupplemented) dams (Table 3). Finally, MCS normalized choline metabolism in the trisomic mice. Ts65Dn adult offspring of supplemented dams exhibited similar (P≥0.068) or higher (P≤0.05) brain choline metabolite enrichments than the unsupplemented 2N mice (Supplemental Tables S2–S4).

Effect of MCS during the perinatal period on overall enrichment of d3- and d6- or d9-choline metabolites in the adult offspring

A main effect of MCS on the metabolic use of d9-choline was detected in liver (+53%, P<0.001), plasma (+34%, P=0.001), and several brain regions (i.e., hippocampus, neocortex, cerebellum, and the rest of the brain, +18 to +22%, P≤0.05), reflecting the higher d9 enrichment in the adult offspring of supplemented (vs. unsupplemented) dams (Fig. 3A). In addition, d9 enrichment in basal forebrain tended to be higher (+16%, P=0.07) among adult offspring born to supplemented (vs. unsupplemented) dams. MCS interacted with genotype to affect d3 and d6 enrichment in liver (P<0.001), plasma (P<0.001), and basal forebrain (P=0.042), with similar trends seen in hippocampus (P=0.1), neocortex (P=0.077), and cerebellum (P=0.075) (Fig. 3B). All of these interactions reflected a significantly greater effect of MCS for the trisomic offspring than their 2N littermates. Specifically, Ts65Dn adult offspring of supplemented (vs. unsupplemented) dams exhibited significantly higher (P≤0.001) d3 and d6 enrichment in liver (+159%), plasma (+147%), basal forebrain (+57%), hippocampus (+51%), neocortex (+57%), and cerebellum (+57%), whereas 2N adult offspring of supplemented (vs. unsupplemented) dams exhibited modestly higher (but significant; P≤0.022) d3 and d6 enrichment in liver (+26%), basal forebrain (+21%), hippocampus (+23%), neocortex (+25%), and cerebellum (+23%) (Fig. 3B). For the metabolic use of endogenously produced d3- and d6-choline in the rest of the brain, a main effect of MCS was detected, with a significantly higher (+36%, P<0.001) d3 and d6 enrichment in the adult offspring of supplemented (vs. unsupplemented) dams (Fig. 3C). Finally, MCS normalized d9-choline metabolite enrichment and improved the d3- and d6-choline metabolite enrichment in trisomic mice. Specifically, in liver, plasma, and all examined brain regions, Ts65Dn adult offspring of supplemented dams exhibited d9-enrichment that was similar (P≥0.08) to the unsupplemented 2N mice, and d3 and d6 enrichment that was significantly higher (P<0.009) than these mice (Supplemental Tables S1–S4).

Figure 3.

Effects of MCS on overall d9-enrichment (A) and overall d3 and d6 enrichment (B, C) in liver, plasma, and brain regions of adult Ts65Dn and 2N offspring born to unsupplemented control vs. MCS dams. MCS did not interact with genotype (P>0.14) for dependent variables shown in A and C; thus the main effect of MCS is presented. Significant (P≤0.042) and borderline significant (P≤0.1) MCS and genotype interactions were detected for the dependent variables shown in B; thus, data were stratified by genotype, and the effect of MCS is presented separately for each genotype (2N-MCS vs. 2N-C; Ts65Dn-MCS vs. Ts65Dn-C). Data are estimated marginal means with 95% confidence intervals. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

DISCUSSION

The study findings show that choline metabolism is altered in trisomic mice, with evidence of preferential partitioning of choline toward the brain; and MCS has lasting effects on the metabolic use of choline in adult offspring, including enhanced de novo production, and utilization, of PEMT-PC molecules in both genotypes, but to a great extent in trisomic offspring.

Choline metabolism is altered in trisomic mice

Ts65Dn (vs. 2N) offspring of unsupplemented dams exhibited substantially lower enrichment of d3- and d6- and d9-choline metabolites in liver and plasma but not in brain tissue. Indeed, brain choline metabolite enrichment was remarkably unaffected by genotype, suggesting that choline is preferentially partitioned to the brain at the expense of other organs in trisomic mice. Redistribution of choline from other organs (e.g., liver, plasma, kidney, and intestines) to the brain has also been demonstrated in a mouse model of choline deprivation (45), suggesting that the trisomy seen in DS increases choline requirements. Based on the present findings, it appears that increasing choline intake by pregnant dams with a trisomic fetus may help normalize their aberrant choline metabolism, which may, in turn, contribute to the observed improvements in cognitive functioning also produced by this maternal dietary intervention.

MCS exerts lasting effects on adult offspring choline metabolism

Both Ts65Dn and 2N offspring of supplemented dams exhibited higher enrichment of d9-choline metabolites in liver and several brain regions, indicating that exposure to higher choline levels during the perinatal period permanently enhances choline uptake and metabolism in these tissues. The programming effect of MCS on d9-choline metabolism was most pronounced in liver (+53% d9-choline metabolite enrichment in offspring of supplemented vs. unsupplemented dams), although also evident in plasma (+34%) and various brain regions (∼+20%) (Fig. 3). The decreasing magnitude of the effects of MCS from liver to plasma to brain suggests that hepatic choline metabolism is most responsive to the programming effect of perinatal choline, and that changes in other organs likely arise from the primary alterations that occurred in liver. This hypothesis is consistent with the central role of liver in supplying extrahepatic tissue with choline and other nutrients (46).

Notably, activity of the hepatic PEMT pathway was 60% higher in Ts65Dn and 2N offspring of choline-supplemented dams. This pathway functions in the de novo biosynthesis of choline (i.e., PC) and the mobilization of hepatic DHA into circulation (43, 47, 48). Specifically, the PEMT pathway (vs. CDP-choline pathway) produces a PC molecule that is enriched in DHA (43, 47, 48) and other long-chain unsaturated fatty acids, which can be incorporated into very low density lipoproteins, released into circulation, and made available to extrahepatic organs (43, 47, 48). Although recycling of PEMT-derived PC to produce PC via the CDP-choline pathway could alter the fatty acid profile of PC, PEMT-PC is mobilized to extrahepatic tissue before extensive recycling occurs (43). Thus, up-regulation of the hepatic PEMT pathway by MCS would be expected to increase both choline and DHA supply to brain and other tissues. Indeed, higher levels of the PEMT-derived d3- and d6-choline metabolites were observed in plasma and several brain regions among both genotypes but to a greater extent in Ts65Dn offspring. The genotype-specific programming effects of MCS may reflect a higher demand for PEMT products (e.g., choline and DHA) among adult Ts65Dn (vs. 2N) offspring, which is consistent with their higher choline requirement. Future attempts at employing d9-choline and stable isotope methodology to study the kinetics of DHA supply from liver to brain should employ a more dynamic pulse-chase approach and include an analysis of the fatty acid composition of the newly synthesized PEMT-PC before recycling of PEMT-PC to produce PC via the CDP-choline pathway occurs.

Enhancement of the hepatic PEMT pathway may beneficially influence cognitive functioning

The long-lasting enhanced supply of choline and DHA to the brain as a result of up-regulation of the hepatic PEMT pathway by MCS may contribute to the improved cognitive functioning produced by this maternal intervention. DHA is abundant among the phospholipids in brain membranes (49) and appears to be particularly important in the development and maintenance of brain mechanisms underlying cognitive functions (50). Administration of nutritional regimens (i.e., supplemental DHA, choline, and uridine) designed to increase PC synthesis and phospholipid-DHA content has been shown to increase PC content in brain membranes, increase dendritic spines, and thus synapses, on hippocampal neurons in animals, enhance neurotransmitter release, and improve cognitive function (reviewed in refs. 49, 51). In addition, the long-lasting elevated supply of choline and DHA as a result of enhanced hepatic PEMT activity in the offspring of choline-supplemented dams may attenuate neurodegeneration, a characteristic of normal aging that is accelerated in DS and AD (7, 52). For example, administering choline to adult mice improves memory in a mouse model of dementia (53) while administering DHA to adult mice reduces amyloid burden in a mouse model of AD (54). Further, provision of choline- and DHA-containing nutritional supplements to older adults with mild AD was shown to improve memory (55). These findings collectively suggest that the lifelong increased activity of the hepatic PEMT pathway produced by MCS and consequent increased provision of PEMT-PC (and its associated DHA) to the brain may mediate some of the cognitive benefits previously reported in adult Ts65Dn and 2N offspring born to choline-supplemented dams (21, 22). However, a limitation of the present study was the lack of data on the fatty acid composition of the PC molecules and other phospholipids (e.g., PE) in liver and brain tissue. These data, along with measurements of PEMT activity (and their correlations with cognitive outcomes), are needed to demonstrate that up-regulation of hepatic PEMT would enhance DHA supply to the brain and contribute to the MCS-induced cognitive benefits. Also requiring clarification is the exact time period (e.g., prenatal or postnatal) required to reap the benefits of MCS delivery and whether the programming effects of MCS on offspring choline metabolism are specific to females.

CONCLUSIONS

Ts65Dn (vs. 2N) mice exhibit alterations in choline metabolism indicative of a higher requirement for this critical bioactive nutrient. MCS increases hepatic PEMT activity in both trisomic and disomic adult offspring and results in higher enrichment of PEMT-derived d3 and d6 metabolites in the brain for both genotypes, but with a more pronounced effect in the Ts65Dn offspring. The increased hepatic supply of PEMT-PC (and its associated fatty acids) to the brain may be one mechanism through which MCS induces lifelong cognitive benefits in the Ts65Dn mouse model of DS and AD, as well as the normal offspring. Translation of MCS to human populations is increasingly warranted based on the collective behavioral, neurochemical, and metabolomic evidence that is being accrued in this highly relevant animal model. In the interim, the present data provide support for advising all pregnant women to consume recommended amounts of choline, especially those carrying a DS fetus.