Late-gestation maternal dietary methyl donor and cofactor supplementation in sheep partially reverses protection against allergic sensitization by IUGR

Abstract

Perinatal exposures are associated with altered risks of childhood allergy. Human studies and our previous work suggest that restricted growth in utero (IUGR) is protective against allergic disease. The mechanisms are not clearly defined, but reduced fetal abundance and altered metabolism of methyl donors are hypothesized as possible underlying mechanisms. Therefore, we examined whether late-gestation maternal dietary methyl donor and cofactor supplementation of the placentally restricted (PR) sheep pregnancy would reverse allergic protection in progeny. Allergic outcomes were compared between progeny from control pregnancies (CON; n = 49), from PR pregnancies without intervention (PR; n = 28), and from PR pregnancies where the dam was fed a methyl donor plus cofactor supplement from day 120 of pregnancy until delivery (PR + Methyl; n = 25). Both PR and PR + Methyl progeny were smaller than CON; supplementation did not alter birth size. PR was protective against cutaneous hypersensitivity responses to ovalbumin (OVA; P < 0.01 in singletons). Cutaneous hypersensitivity responses to OVA in PR + Methyl progeny were intermediate to and not different from the responses of CON and PR sheep. Cutaneous hypersensitivity responses to house dust mites did not differ between treatments. In singleton progeny, upper dermal mast cell density was greater in PR + Methyl than in PR or CON (each P < 0.05). The differences in the cutaneous allergic response were not explained by treatment effects on circulating immune cells or antibodies. Our results suggest that mechanisms underlying in utero programming of allergic susceptibility by IUGR and methyl donor availability may differ and imply that late-gestation methyl donor supplementation may increase allergy risk.

Keywords: animal models; fetal growth; folic acid; intrauterine growth restriction; mast cells; methyl donors.

Fig. 1.

Circulating 1-carbon metabolite concentrations in ewes and lambs on the day of birth in control (CON; open bars and open circles), placentally restricted (PR; black bars and black circles), and PR with maternal dietary methyl donor and cofactor supplementation (PR + M; gray bars and gray-filled circles) and in singleton (black and gray-hatched bars) and twin (black and white hatched bars) litters and relationships between maternal and neonatal abundance. A: plasma folate (ewes: CON, n = 15; PR, n = 9; PR + M, n = 9; lambs: CON, n = 24; PR, n = 11; PR + M, n = 9). B: plasma methionine (ewes: CON, n = 14; PR, n = 9; PR + M, n = 10; lambs: CON, n = 24; PR, n = 11; PR + M, n = 11). C: plasma S-adenosyl-methionine (ewes: CON, n = 14; PR, n = 9; PR + M, n = 10; lambs: CON, n = 24; PR, n = 11; PR + M, n = 11). Plasma concentrations of methionine and S-adenosyl-methionine represent relative ionic abundance obtained by liquid chromatography-mass spectrometry analysis. Progeny outcomes were analyzed using a generalized linear mixed models framework that examined the effects of treatment (CON, PR, PR + M) and litter size (singleton birth vs. twin birth), treating the dam as the experimental unit and sibling data as repeated measurements on each dam. No significant interactions were evident between treatment and litter size, and data are shown as estimated means ± SE for each main factor. Where treatment effects or trends were apparent (P < 0.1), we compared means for each treatment by the least significant difference method, as described under statistical analyses. Within treatments (a and b) or litter size groups (c and d) in ewes or lambs, groups that do not share a common letter differ (P < 0.05). Relationships between ewe and neonatal lamb concentrations for each metabolite for all ewe-lamb pairs, including twins, were assessed by Pearson’s correlation; overall relationships and 95% confidence intervals are shown at right, whereas relationships within treatment groups are described in results.

Fig. 2.

Proportion of positive (cross-hatched portion, skin wheel diameter ≥3 mm) and negative (solid colored portion, skin wheel diameter <3 mm) responders at 24 h after intradermal challenge with ovalbumin in control (CON; white portion), placentally restricted (PR; black portion), and PR with maternal dietary methyl donor and cofactor supplementation (PR + M; gray portion) after a sensitization protocol. A: overall cohort (CON: n = 49; PR: n = 28; PR + M: n = 25). B: singleton-birth sheep only (CON: n = 9; PR: n = 18; PR + M: n = 13). Binary outcomes (responders and nonresponders) were analyzed using a generalized linear mixed models framework that examined the effects of treatment (CON, PR, PR + Methyl), litter size (singleton birth vs. multiple birth), and sex difference, treating the dam as the experimental unit and sibling data as repeated measurements on each dam, assuming a binomial distribution and utilizing a logit link function. Where treatment effects or trends were apparent (P < 0.1), we compared means for each treatment by the least significant difference method, as described in Statistical Analysis. **P < 0.01, differences between groups.

Fig. 3.

A: upper dermis of skin sections from adult sheep were stained with toluidine blue; mast cells (indicated by arrows) stain metachromatically purple. Scale bar is 100 µm in length. B: upper dermis mast cell density in singleton-birth and multiple-birth male (open bars) and female (closed bars) sheep in control (CON; n = 49), placentally restricted (PR, n = 28), and PR with maternal dietary methyl donor and cofactor supplementation (PR + M; n = 25). Outcomes were analyzed using a generalized linear mixed models framework that examined the effects of treatment (CON, PR, PR + Methyl), litter size (singleton birth vs. multiple birth), and sex difference, treating the dam as the experimental unit and sibling data as repeated measurements on each dam. Where treatment effects or trends were apparent (P < 0.1), we compared means for each treatment by the least significant difference method, as described in Statistical Analysis. Values are estimated means ± SE. *P < 0.05; **P < 0.01; ***P < 0.005.