Postnatal adaptations of phosphatidylcholine metabolism in extremely preterm infants: implications for choline and PUFA metabolism

Abstract

Background: Lipid metabolism in pregnancy delivers PUFAs from maternal liver to the developing fetus. The transition at birth to diets less enriched in PUFA is especially challenging for immature, extremely preterm infants who are typically supported by total parenteral nutrition.

Objective: The aim was to characterize phosphatidylcholine (PC) and choline metabolism in preterm infants and demonstrate the molecular specificity of PC synthesis by the immature preterm liver in vivo.

Methods: This MS-based lipidomic study quantified the postnatal adaptations to plasma PC molecular composition in 31 preterm infants <28 weeks' gestational age. Activities of the cytidine diphosphocholine (CDP-choline) and phosphatidylethanolamine-N-methyltransferase (PEMT) pathways for PC synthesis were assessed from incorporations of deuterated methyl-D9-choline chloride.

Results: The concentration of plasma PC in these infants increased postnatally from median values of 481 (IQR: 387-798) µM at enrollment to 1046 (IQR: 616-1220) µM 5 d later (P < 0.001). Direct incorporation of methyl-D9-choline demonstrated that this transition was driven by an active CDP-choline pathway that synthesized PC enriched in species containing oleic and linoleic acids. A second infusion of methyl-D9-choline chloride at day 5 clearly indicated continued activity of this pathway. Oxidation of D9-choline through D9-betaine resulted in the transfer of 1 deuterated methyl group to S-adenosylmethionine. A very low subsequent transfer of this labeled methyl group to D3-PC indicated that liver PEMT activity was essentially inactive in these infants.

Conclusions: This study demonstrated that the preterm infant liver soon after birth, and by extension the fetal liver, was metabolically active in lipoprotein metabolism. The low PEMT activity, which is the only pathway for endogenous choline synthesis and is responsible for hormonally regulated export of PUFAs from adult liver, strongly supports increased supplementation of preterm parenteral nutrition with both choline and PUFAs.

Keywords: CDP-choline pathway: PEMT pathway; plasma phosphatidylcholine; preterm infants; stable isotopes.

FIGURE 1

Metabolic pathway of intravenously administered choline and its methyl groups using deuterium-labeled methyl-D₉-choline. Methyl groups labeled with deuterium are represented by the black dots; the unlabeled methyl groups are represented by the white dots. The methyl-D₉-choline can enter the CDP-choline pathway (*) to produce D₉-PC. Alternatively, it can be oxidized to D₉-betaine, which then donates a single methyl group to homocysteine, producing D₆-DMG and D₃-methionine. D₃-methionine serves as a precursor for D₃-SAMe, which can be used by PEMT (‡) to sequentially methylate PE, forming D₃-PC as well as smaller amounts of D₆-PC. A fraction of D₃-choline or D₆-choline is released by hydrolysis of PEMT-derived D₃-PC or D₆-PC in the liver and recycled back to D₃-PC or D₆-PC by the CDP-choline pathway. CDP-choline, cytidine diphosphocholine; DMG, dimethylglycine; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PEMT, phosphatidylethanolamine N-methyltransferase; S-AdHo, S-adenosylhomocysteine; SAMe, S-adenosylmethionine.

FIGURE 2

Concentration of plasma PC in preterm infants. (A) PC concentration increased with time from recruitment and exhibited considerable variation. Results are shown as a box-and-whisker plot with medians, IQRs. The symbols above the box-and-whisker plots represent outlier values. Each symbol refers to 1 of 3 of the same infants who had high concentrations of plasma PC. Significance was assessed using the Mann-Whitney test. (B) PC concentration from time of recruitment for individual infants, expressed relative to the initial value at t = 0 h. (C) Concentrations of selective PC molecular species over the first 5 d of the study. Results are expressed as means ± SEMs. Numbers of infants decreased at each time point from 31 at t = 0 to 23 at t = 120 h. PC, phosphatidylcholine.

FIGURE 3

PC molecular species composition of preterm infant plasma. Results are presented at recruitment and expressed as a percentage of total PC (t = 0 h, n = 31, open bars) and at 5 d (t = 120 h, n = 23, closed bars). Values are means ± SDs. *P < 0.001 (t test). LPC, lysophosphatidylcholine; PC, phosphatidylcholine.

FIGURE 4

Incorporation of stable isotope–labeled choline into plasma PC in preterm infants. (A) The fractional enrichment of methyl-D₉-PC (solid line, mean ± SEM) synthesized by the CDP-choline pathway and of methyl-D₃-PC synthesized by the PEMT pathway (dashed line, mean with SEM smaller than symbol). There were 31 infants at t = 0, and numbers of infants at each time point are indicated on the figure. (B) The same results as in panel A after normalization to the initial concentration value using the variation in PC concentration with time detailed in Figure 2B. CDP-choline, cytidine diphosphocholine; PC, phosphatidylcholine; PEMT, phosphatidylethanolamine N-methyltransferase.

FIGURE 5

Enrichment of methyl-D₉-choline into individual molecular species of plasma PC. Results are presented for incorporations into methyl-D₉-PC (A, C) and methyl-D₃-PC (B, D). The results in panels A and B were expressed as percentage labeled:labeled + unlabeled PC. Panels C and D present the same results after normalization to initial concentrations of individual PC molecular species at t = 0 h. Results are presented as means ± SEMs; numbers at each time point were the same as in Figure 4A. PC, phosphatidylcholine.

FIGURE 6

The effect of disease severity and immaturity on label incorporation into PC molecular species by the PEMT pathway. Combined enrichments of the polyunsaturated species PC 38:4, PC 38:5, PC 38:6, and PC 40:6 were calculated in panel A for infants with more severe disease who received a second dose of methyl-D₉-choline at 120 h compared with infants who received a single dose of label (mean ± SEM; *P < 0.05, t test). Initially, 14 infants received the single dose and 13 infants received 2 doses. Corresponding numbers at 120 h were 6 and 11 infants. Infants with too few samples to calculate label incorporation (n = 4) were excluded. The association of PEMT, calculated from methyl-D₃ label enrichment at 24 and 48 h, is shown for gestational age (B) and birth weight (C). PC, phosphatidylcholine; PEMT, phosphatidylethanolamine N-methyltransferase.

FIGURE 7

Fractional rate of plasma PC synthesis by the PEMT pathway. (A) The enrichment of hepatic SAMe was determined by MIDA calculation based on the relative incorporations of the methyl-D₃ and methyl-D₆ labels into PC (mean ± SEM, n = 27). (B) Rate of fractional synthesis by the PEMT pathway after correcting enrichment of methyl-D₃-PC for enrichment of SAMe at each time point (% total PC). MIDA, mass isotopomer distribution analysis; PC, phosphatidylcholine; PEMT, phosphatidylethanolamine N-methyltransferase; SAMe, S-adenosylmethionine.

FIGURE 8

Synthesis of plasma LPC. (A) Composition of individual molecular species of plasma LPC (% total PC + LPC; mean ± SEM). (B) Enrichments of methyl-D₉ (solid line) and methyl-D₃ (dashed line) labels in plasma LPC species normalized to their initial concentrations at t = 0 h (mean ± SEM). (C) Enrichment of methyl-D₉-choline in LPC species over the first 5 d after recruitment. Numbers at each time point were the same as in Figure 4A. LPC, lysophosphatidylcholine; PC, phosphatidylcholine.