Transgenerational rescue of a genetic defect in long-term potentiation and memory formation by juvenile enrichment

Abstract

The idea that qualities acquired from experience can be transmitted to future offspring has long been considered incompatible with current understanding of genetics. However, the recent documentation of non-Mendelian transgenerational inheritance makes such a "Lamarckian"-like phenomenon more plausible. Here, we demonstrate that exposure of 15-d-old mice to 2 weeks of an enriched environment (EE), that includes exposure to novel objects, elevated social interactions and voluntary exercise, enhances long-term potentiation (LTP) not only in these enriched mice but also in their future offspring through early adolescence, even if the offspring never experience EE. In both generations, LTP induction is augmented by a newly appearing cAMP/p38 MAP kinase-dependent signaling cascade. Strikingly, defective LTP and contextual fear conditioning memory normally associated with ras-grf knock-out mice are both masked in the offspring of enriched mutant parents. The transgenerational transmission of this effect occurs from the enriched mother to her offspring during embryogenesis. If a similar phenomenon occurs in humans, the effectiveness of one's memory during adolescence, particularly in those with defective cell signaling mechanisms that control memory, can be influenced by environmental stimulation experienced by one's mother during her youth.

Figure 1.

Long-lasting effects of juvenile enrichment. WT (a) or ras-grf knock-out (b) mice (1 month old) were exposed to an enriched environment for 2 weeks and then placed back into conventional environment for various times. LTP was then assayed. Experiments represent mean ± SEM from at least six brain slices from at least two independent experiments. c, A timeline that explains these experiments (F0) and experiments described in Figure 2 (F1), and Figures 2–5 (F1 and F2). The bars represent the life cycle of each generation relative to each other and when animals were tested (arrows).

Figure 2.

Adolescent offspring of enriched juvenile ras-grf knock-out mice display normal LTP even if they are raised in conventional environment. a, LTP assays on hippocampal slices from 1-month-old ras-grf1/grf2 double knock-out mice exposed to either a conventional environment (black squares) or 2 weeks of enriched environment (red circles). n is number of slices and N is number of animals used. b, LTP assays of 1-month-old F1 offspring of either enriched (red circles) or unenriched (black squares) ras-grf knock-out mice. c, LTP assays on hippocampal slices from 1-month-old WT mice exposed to either a conventional environment (black squares) or 2 weeks of enriched environment (red circles). d, LTP assays of 1-month-old F1 offspring of either enriched (red circles) or unenriched (black squares) WT mice. Inset traces show a typical fEPSP from each group, gray traces are taken from 5 min before TBS, the colored traces are taken 50 min after TBS. Calibration: 1 mV, 10 ms. e, Immunoblot of brain slices demonstrating GRF1 and GRF 2 expression in WT but not F1 generation ras-grf1/2 double knock-out mice (mutant offspring) from enriched parents. The data represent the average and SEM from at least three independent experiments.

Figure 3.

Adolescent offspring of enriched mice display the same novel LTP inducing signaling pathway as their enriched parents. a, LTP assays on hippocampal slices from 1-month-old offspring of enriched ras-grf1/grf2 double knock-out mice pretreated with buffer (black squares), NMDA receptor inhibitor (AP-5, 100 μm) (red circles), or cAMP inhibitor (Rp-cAMPS, 100 μm) (green triangles). b, Same as in a except that samples were pretreated with vehicle (black squares), p38 inhibitor (SB203580, 5 μm), or inhibitor of Erk signaling (U0126, 20 μm) (green triangles). c, Same as in a except that offspring of enriched WT mice were studied and samples were pretreated with vehicle (black squares), p38 inhibitor (red circles), or cAMP inhibitor (green triangles). d, Same as in c except that offspring of enriched WT mice were studied and data were pooled from multiple experiments. The data represent the average and SEM from the indicated number of samples.

Figure 4.

Mechanism of transmission of EE effects across generations. a, LTP assays on hippocampal slices from 1-month-old offspring of enriched female and unenriched male (black squares) or enriched male and unenriched female (red circles) ras-grf1/grf2 double knock-out mice. b, LTP assays on hippocampal slices from the offspring of enriched ras-grf knock-out mice but raised by unenriched foster mothers (EE-Biologic-black squares) or offspring of unenriched ras-grf knock-out mice raised by enriched foster mothers (EE-Foster-red circles). c, LTP assays on 1-month-old F1-generation offspring (black squares), or F2 second-generation offspring (red circles) of enriched ras-grf knock-out mice. Samples from 1-month-old unenriched ras-grf knock-out mice (green triangles) are also shown for comparison. d, LTP assays were performed on hippocampal slices from enriched ras-grf knock-out mice (red bars) and F1 offspring of enriched ras-grf knock-out mice (green bars) at 1 month of age. Mice were then returned to a conventional environment and assayed at 2 months of age and 3 months of age. Data are the average fold increase in fEPSP slopes from three experiments (Fig. 1C).

Figure 5.

Effect of enrichment on contextual conditioning memory. a, ras-grf1/2 double knock-out mice have defective contextual conditioning memory. One-month-old wild-type mice (□) (n = 17) and ras-grf1/2 double knock-out mice (■) (n = 15) (F0) were subjected to one-trial paradigm training. There was no significant difference in baseline freezing before footshock (pre) or immediately after (post). A retention test was performed 24 h after training. The data represent the average and SEM from the indicated number of samples. The freezing response in ras-grf1/2 double knock-out mice was significantly lower than that in wild-type mice (p < 0.05, Mann–Whitney U test). b, Enrichment rescues deficient contextual conditioning memory in ras-grf1/2 double knock-out mice. One-month-old naive ras-grf1/2 double knock-out mice control (□) (n = 12) and enriched ras-grf1/2 double knock-out mice (■) (n = 12) were subjected to one-trial paradigm training as described in a. The freezing response in enriched ras-grf1/2 double knock-out mice was significantly higher than that in naive ras-grf1/2 double knock-out mice (p < 0.05, Mann–Whitney U test). c, Defective contextual conditioning memory is partially rescued in 1-month-old offspring of enriched juvenile ras-grf1/2 double knock-out mice. One-month-old offspring from two litters of either naive ras-grf1/2 double knock-out mice (□) (n = 15) or enriched ras-grf1/2 double knock-out mice (■) (n = 16) were subjected to one-trial paradigm training as in a. The freezing response in the offspring of enriched ras-grf1/2 double knock-out mice was significantly higher than that in naive ras-grf1/2 double knock-out mice (p < 0.05, Mann–Whitney U test). d, Defective contextual fear memory is not rescued in 3-month-old F1 offspring of enriched juvenile ras-grf1/2 knock-out mice. Experiments were performed as in a except that offspring of enriched knock-out mice were reared in a standard environment for 3 months before being tested for fear conditioning response (control, n = 12) (EE, n = 15). e, Defective contextual conditioning memory is not rescued in the F2 offspring of enriched juvenile ras-grf1/2 knock-out mice. Experiments were performed as in a except that 1-month-old F2 offspring of enriched knock-out mice were used (control, n = 11) (EE, n = 13).