Transgenerational epigenetic heritability for growth, body composition, and reproductive traits in Landrace pigs

Abstract

Epigenetics is an important source of variation in complex traits that is not due to changes in DNA sequences, and is dependent on the environment the individuals are exposed to. Therefore, we aimed to estimate transgenerational epigenetic heritability, percentage of resetting epigenetic marks, genetic parameters, and predicting breeding values using genetic and epigenetic models for growth, body composition, and reproductive traits in Landrace pigs using routinely recorded datasets. Birth and weaning weight, backfat thickness, total number of piglets born, and number of piglets born alive (BW, WW, BF, TNB, and NBA, respectively) were investigated. Models including epigenetic effects had a similar or better fit than solely genetic models. Including genomic information in epigenetic models resulted in large changes in the variance component estimates. Transgenerational epigenetic heritability estimates ranged between 0.042 (NBA) to 0.336 (BF). The reset coefficient estimates for epigenetic marks were between 80% and 90%. Heritability estimates for the direct additive and maternal genetic effects ranged between 0.040 (BW) to 0.502 (BF) and 0.034 (BF) to 0.134 (BW), respectively. Repeatability of the reproductive traits ranged between 0.098 (NBA) to 0.148 (TNB). Prediction accuracies, bias, and dispersion of breeding values ranged between 0.199 (BW) to 0.443 (BF), -0.080 (WW) to 0.034 (NBA), and -0.134 (WW) to 0.131 (TNB), respectively, with no substantial differences between genetic and epigenetic models. Transgenerational epigenetic heritability estimates are moderate for growth and body composition and low for reproductive traits in North American Landrace pigs. Fitting epigenetic effects in genetic models did not impact the prediction of breeding values.

Keywords: ebv; epigenetics; genetic parameters; genomics; swine; variance components.

FIGURE 1

Correlation among posterior samples of the parameters used in the genetic, epigenetic, genomic, and epigenetic models including genomic information (Epi_genomic) for birth weight (BW), weaning weight (WW), backfat (BF), total number of piglets born (TNB), and number of piglets born alive (NBA) in Landrace pigs. v_add, v_mat, v_c, v_pe, v_epi, v_res, p2, cov_add_mat = additive genetic, maternal genetic, common environment, permanent environment, transgenerational epigenetic, residual, and phenotypic variances, and covariance between additive genetic and maternal genetic effects, respectively. The heritability and ratios had the same pattern of their respective variance components and were omitted for simplicity. Blank squares mean non-significant correlation coefficient at 5% of probability.

FIGURE 2

Correlation among solutions from mixed model equations using a genetic and epigenetic model for the birth weight, weaning weight, backfat, total number of piglets born, and number of piglets born alive in Landrace pigs. gen_add_bw, gen_mat_bw, epi_add_bw, epi_mat_bw, epi_epi_bw = solutions for the direct additive genetic effect from genetic model, solutions for the maternal genetic effect from genetic model, solutions for the direct additive genetic effect from epigenetic model, solutions for the maternal genetic effect from epigenetic model in the birth weight, respectively; same pattern was employed for the other traits changing the trait suffix for ww, bf, tnb, and nba, for weaning weight, back fat, total number born, and number born alive, respectively, the only exception was for total number born and number born alive, that did not have maternal genetic effects and presented permanent environment effect, called as gen_pe_tnb, epi_pe_tnb for the solutions for the permanent environment effects for genetic and epigenetic models for total number born, respectively, same pattern was applied for number born alive.

FIGURE 3

Distribution of the solutions and difference among solutions between the genetic and epigenetic models for the random effects used to analyze birth weight (BW), weaning weight (WW), backfat (BF), total number of piglets born (TNB), and number of piglets born alive (NBA) in Landrace pigs. Add, Mat, PE, Gen, Epi, Diff, ***, NS = direct additive genetic solution, maternal genetic solution, permanent environment solution, genetic model, epigenetic model, difference between genetic and epigenetic solutions, significant at 0.1% by a paired t-test, and not significant (P > 0.05), respectively.

FIGURE 4

Prediction accuracy, bias, and dispersion of the estimated breeding values of focal animals for birth weight (birth_weight), weaning weight (weaning_weight), backfat (back_fat), total number of piglets born (total_number_born), and number of piglets born alive (number_born_alive) in Landrace pigs. Epi, Gen, = epigenetic model and genetic model, respectively.