Does genomic imprinting contribute to valproic acid teratogenicity?

Abstract

Reciprocal outcrosses and backcrosses were made between strains of mice with different susceptibilities to valproic acid (VPA) teratogenicity. Relatively resistant C57BL/6J (C) and more susceptible SWV (S) strains of mice produced F1 hybrids in which the female parent was C and the male parent was S (CS-F1) as well as the reciprocal with S dams and C sires (SC-F1). Each was backcrossed to each strain, producing 8 types of backcross matings: CS x C, SC x C, CS x S, SC x S; C x CS, C x SC, S x CS, S x SC (for all matings dams are listed first). At 8d:12 +/- 5h of gestation, a teratogenic dose, 600 mg/kg, of aqueous VPA was injected ip into the dams. Fetuses were examined on gestation day (gd) 18 for abnormality, mortality, litter size, and weight. Genomic imprinting (imprinting) is a phenomenon at least in part involving hyper- or hypomethylation of bases in DNA, which is believed to determine whether or not the imprinted gene will be expressed. Imprinting has been reported to occur differentially in the male and female for a number of gene loci. Thus, in crosses between strains with differing susceptibility to VPA, if imprinting is occurring, the susceptibility of a fetus might be predicted to be disproportionately influenced by susceptibility of its grandparents. Significant differences in frequency (%) of occurrence of exencephaly in progeny of all backcrosses with F1 dams consistent with those expected for imprinting were found in the present study (CS-F1x C = 21.8 +/- 3.9%, SC-F1x C = 10.8 +/- 3.2%, P < 0.03; CS-F1x S = 14.8 +/- 3.1%, SC-F1x S = 6.3 +/- 2.3%, P < 0.03). SWV dams revealed the same pattern (S x SC-F1 = 50.0 +/- 8.3%, S x CS-F1 = 37.1 +/- 4.7%, P < 0.04). Differences in prenatal mortality also consistent with genomic imprinting occurred in backcrosses with pure-line SWV dams (S x SC = 64.4 +/- 8.0%, S x CS = 30.5 +/- 4.5%, P < 0.001). Fetal weight was reduced in a manner consistent with imprinting in backcrosses involving SWV (S x SC = 0.50 +/- 0.18 g, S x CS = 0.96 +/- 0.05, P < 0.01). Three of four of the parameters investigated showed differences in some of the backcrosses of reciprocal F1's consistent with those expected if genomic imprinting were occurring.