Epigenetic Transmission of the Impact of Early Stress Across Generations

Experimental design and maternal care provided to first- and second-generation (F1 and F2) maternal separation with unpredictable maternal stress (MSUS) and control pups. (A) Experimental design used to study the impact of MSUS on behavior and its transmission across generations. C57Bl6/J F0 females (black) bred to C57Bl6/J males were allowed to raise their offspring in normal conditions (control; left, blue) or were subjected to MSUS from postnatal day (PND) 1 to 14 (right, orange), and maternal care was observed. The F1 progeny (females and males) was weaned, raised normally until adulthood, then behaviorally tested. Following testing, adult F1 control (blue, left) and F1 MSUS (orange, right) males were bred to C57Bl6/J females (black), and F2 offspring were raised in normal conditions (no maternal separation or maternal stress). F2 control and MSUS males were bred to C57Bl6/J females, and F3 offspring were raised in normal conditions. The dotted arrow symbolizes separation of dams from pups. Poor maternal care provided to F1 MSUS pups did not result in abnormal weight during postnatal development. Dams undergoing MSUS (n = 7) spent (B) less time actively nursing [F(1,13) = 7.25, p < .05] and (C) more time off nest [F(1,13) = 7.69, p < .05] than control dams (n = 8) during the first week postdelivery (PND 1–7). Average of maternal care scoring across three 30-min sessions per day. Similar level of maternal care was provided to F2 MSUS and control mice derived from F1 MSUS males. Daily scoring of maternal care from PND 1 to 7 (three 30-min sessions/day) showing similar (D) level of active nursing [arched-back nursing (ABN) and ABN associated with licking-grooming (ABN + ABN-LG), in percent; t(17) = 1.10, ns] and (E) time off-nest [t(17) = 1.02, ns; control, n = 10; MSUS, n = 9]. Maternal care provided to F2 pups was not monitored from PND 8–14 because it had not changed from PND 1–7. *p < .05 as indicated by Fisher's Protected Least Significant Difference following two-way analysis of variance.

Depressive-like behaviors in first-generation (F1) males and second- and third-generation (F2 and F3) offspring. In the forced swim test, (A) there was increased time spent floating in F1 maternal separation with unpredictable maternal stress (MSUS; n = 29) vs. control males [control, n = 14; t(42) = 2.25, p < .05]; (B) F2 MSUS females (n = 31) compared with F2 control females (n = 28), but not F2 MSUS males (n = 29) compared with F2 control males [n = 30; F(1,114) = 6.95, p < .01]; and (C) in F3 MSUS males (n = 22) but not females (n = 18) compared with controls [males, n = 20; females, n = 19; F(1,75) = 6.09, p < .05]. Lower sucrose intake normalized to water intake over 4 days in (D) F1 MSUS (n = 20) vs. F1 control males [control, n = 18; t(36) = 2.22, p < .05] but not in (E) F2 MSUS vs. F2 control mice [F2 control males, n = 13; F2 MSUS males, n = 16; F2 control females, n = 16; F2 MSUS females, n = 16; F(1,57) = 3.07, .05 < p < .1] or (F) F3 MSUS and control males or females [F3 control males, n = 18; F3 MSUS males, n = 18; F3 control females, n = 19; F3 MSUS females, n = 19; F(1,70) = .36, ns]. *p < .05 as indicated by unpaired t test or Fisher's Protected Least Significant Difference following two-way analysis of variance.

Reversal of depressive-like behaviors in second-generation (F2) maternal separation with unpredictable maternal stress (MSUS) mice with acute and chronic antidepressant treatment. In the forced swim test, increased time spent floating in saline-treated female F2 MSUS (n = 17) vs. F2 control (n = 14) was reversed by both (A) acute (1 day) and (C) chronic (14 day) treatment with desipramine [Desip.; F2 MSUS n = 14, F2 control, n = 11; (A) F(1,52) = 2.93, p < .05; (C) F(1,47) = 3.22, p < .05]. In the tail suspension test, increased time spent immobile in saline-treated female F2 MSUS (n = 16) vs. F2 control (n = 14) was reversed by both (B) acute and (D) chronic treatment with desipramine [F2 MSUS, n = 14, F2 control, n = 11; (B) F(1,50) = 5.93, p < .001; (D) F(1,48) = 7.29, p < .001]. *p < .05, MSUS vs. control within generation/drug treatment; ^#p < .05, saline vs. desipramine within MSUS or control as indicated by Fisher's Protected Least Significant Difference post hoc tests.

Altered behavioral response in first-generation (F1) males and second- and third-generation (F2 and F3) offspring. Reduced latency to enter unfamiliar areas in (A) F1 maternal separation with unpredictable maternal stress (MSUS) males [n = 41; t(64) = 2.60, p < .05], (B) F2 MSUS females (n = 15) but not F2 MSUS males [n = 29; F2, F(1,84) = 4.74, p < .05], and (C) F3 MSUS females (n = 16), but not F3 MSUS males (n = 18), compared with controls [F1 control males, n = 25; F2 control females, n = 15; F2 control males, n = 29; F3 control females, n = 18; F3 control males, n = 19; F3, F(1,67) = 6.94, p < .05] in the free exploratory paradigm. There was a trend for reduced latency to enter the center of the open field in (D) F1 MSUS males (n = 38; t(61) = 1.73, .05 < p < .1], significant reduction in (E) F2 MSUS females (n = 15) but not in F2 MSUS males (n = 29; F(1,83) = 4.88, p < .05], and (F) in F3 MSUS females (n = 17) but not F3 MSUS males (n = 19) compared with controls [F1 controls males, n = 28; F2 control females, n = 15; F2 control males, n = 28; F3 control females, n = 20; F3 control males, n = 18; F3, F(1,35) = 5.37, p < .05]. ^#.05 < p < .1; *p < .05; **p < .01, as indicated by unpaired t test or Fisher's Protected Least Significant Difference following two-way analysis of variance.

Altered methylation of MeCP2, CB1, and the CRFR2 CpG island in the first-generation (F1) germline and second-generation (F2) brain and decreased mRNA expression in F2 brain. Schematic representation of (A) MeCP2, (E) CB1, and (I) CRFR2 genes showing the CpG island and transcription initiation site. Base-bp annotations are relative to the location of the initiation site. Target sequences used for pyrosequencing to quantify methylation are represented as a blue line. (B) Reduced MeCP2 mRNA expression in F2 MSUS (n = 7) compared with F2 control (n = 5) brain [t(10) = 2.80, p < .05]. Increased methylation of the MeCP2 CpG island in (C) sperm of F1 MSUS males [F(1,7) = 6.19, p < .05] and (D) brain of F2 MSUS females [F(1,6) = 13.31, p = .01; sperm, n = 4–5; brain, n = 4]. (F) Reduced CB1 mRNA expression in F2 MSUS (n = 8) compared with F2 control (n = 8) brain [t(14) = 2.21, p < .05]. Increased methylation of the CB1 CpG island in (G) sperm of F1 MSUS males [F(1,9) = 15.57, p < .01] and (H) brain of F2 MSUS females [F(10,40) = 5.26, p < .01; sperm, n = 5–6, brain, n = 3]. (J) Reduced CRFR2 mRNA expression in F2 MSUS (n = 11) compared with F2 control mice [n = 12; t(21) = 2.31, p < .05]. Reduced methylation of the CRFR2 CpG island in (K) sperm of F1 MSUS males (n = 6) compared with F1 control (n = 6) mice [F(1,10) = 5.31, p < .05] and (L) brain in F2 MSUS females [n = 4, F(1,6) = 20.66, p < .05]. ⁺ 05 < p < .1; * p < .05; ** p < .01; *** p < .001, as indicated by Fisher's Protected Least Significant Difference following repeated-measures analysis of variance.