Early-life nicotine or cotinine exposure produces long-lasting sleep alterations and downregulation of hippocampal corticosteroid receptors in adult mice

Abstract

Early-life exposure to environmental toxins like tobacco can permanently re-program body structure and function. Here, we investigated the long-term effects on mouse adult sleep phenotype exerted by early-life exposure to nicotine or to its principal metabolite, cotinine. Moreover, we investigated whether these effects occurred together with a reprogramming of the activity of the hippocampus, a key structure to coordinate the hormonal stress response. Adult male mice born from dams subjected to nicotine (NIC), cotinine (COT) or vehicle (CTRL) treatment in drinking water were implanted with electrodes for sleep recordings. NIC and COT mice spent significantly more time awake than CTRL mice at the transition between the rest (light) and the activity (dark) period. NIC and COT mice showed hippocampal glucocorticoid receptor (GR) downregulation compared to CTRL mice, and NIC mice also showed hippocampal mineralocorticoid receptor downregulation. Hippocampal GR expression significantly and inversely correlated with the amount of wakefulness at the light-to-dark transition, while no changes in DNA methylation were found. We demonstrated that early-life exposure to nicotine (and cotinine) concomitantly entails long-lasting reprogramming of hippocampal activity and sleep phenotype suggesting that the adult sleep phenotype may be modulated by events that occurred during that critical period of life.

Figure 1

Graphical representation of experimental design. The picture shows the scheme of the experimental protocol used in this study (cf. Methods). C57Bl/6 J dams (3–4/cage) drank water laced with 2% saccharin, or 2% saccharin + 100 µg/mL nicotine freebase or 2% saccharin + 10 µg/mL cotinine for 3 weeks prior to and 6 weeks after mating. Pups never drank these solutions. Adult (≈ 17 weeks old) male mice underwent surgery to implant electrodes for electroencephalographic (EEG) and electromyographic (EMG) signals. After at least 2 weeks of recovery, basal sleep phenotype was assessed by recording EEG/EMG signals for 48 h. After that, mice were split in 2 subgroups to evaluate either the sleep homeostasis (after 6 h of sleep deprivation) or the hippocampal expression and methylation of corticosteroid receptor genes.

Figure 2

Sleep fragmentation index. Values of sleep fragmentation index calculated as the n° of awakenings/total sleep time (in hours) during 48 h of baseline recordings and 18 h of recordings during recovery after sleep deprivation (SD) by gentle handling for 6 h. Values are shown for adult male mice perinatally exposed to nicotine (NIC, n = 15), cotinine (COT, n = 14) or just the vehicle (CTRL, n = 10). Values after sleep deprivation were calculated on 10 NIC, 8 COT and 6 CTRL mice. n.s., not significant.

Figure 3

Day-night hourly profiles of the time spent in wakefulness during baseline recordings. Panel (a) shows the 24-h hourly profile of the time spent in wakefulness (W) during baseline recordings for adult male mice perinatally exposed to nicotine (NIC, n = 15), cotinine (COT, n = 14) or just the vehicle (CTRL, n = 10). Mice were left undisturbed with lights on at 9.00 a.m. (Zeitgeber Time 0, ZT0) and lights off at 9.00 p.m. (ZT12). Panel (b) shows the amount of W in the 4 h surrounding the photoperiod transitions. In particular, values for light to dark (LtoD, ZT10–ZT14) and dark to light (DtoL, ZT22–ZT2) transitions are shown. *p (adjusted for multiple comparisons) < 0.05; n.s., not significant.

Figure 4

Electroencephalogram analysis in sleep during baseline recordings. Panel (a) shows electroencephalographic (EEG) power spectral density in baseline recordings during non-rapid-eye-movement sleep (NREMS) expressed as a percentage of total EEG spectral power. Panel (b) shows power in the delta frequency range (1–4 Hz, EEG slow-wave activity, SWA) during NREMS in baseline conditions. EEG SWA was normalized to values in the last 4 h of the light period. Panel (c) shows EEG power spectral density in baseline recordings during rapid-eye-movement sleep (REMS) expressed as a percentage of total EEG spectral power. Panel (d) shows a magnification of the orange square in panel (c). Panel (e) shows the values of the EEG peak during REMS. All data refer to adult male mice perinatally exposed to nicotine (NIC, n = 14), cotinine (COT, n = 14) or just the vehicle (CTRL, n = 11). *p (adjusted for multiple comparisons) < 0.05; n.s., not significant.

Figure 5

Corticosteroid receptor gene expression in hippocampus. Panel (a, b) show the expression of hippocampal GR (Nr3c1) and MR (Nr3c2) genes relative to the expression of glyceraldehyde-3-phosphate dehydrogenase (Gapdh), used as reference gene. Panel (c) shows the expression ratio of MR and GR genes. Panels (d, e) show the correlation between the relative hippocampal GR (panel d) or MR (panel e) expression (x-axis) and the amount of wakefulness (W) at light to dark (LtoD) transition (y-axis) considering all the mice of the present experiment. Continuous and dotted lines indicate the regression line and the 95% confidence, respectively. All data refer to adult male mice perinatally exposed to nicotine (NIC, n = 5), cotinine (COT, n = 6) or just the vehicle (CTRL, n = 6). *p (adjusted for multiple comparisons) < 0.05; **p (adjusted for multiple comparisons) < 0.01; n.s., not significant.

Figure 6

DNA methylation analysis of hippocampal corticosteroid receptors. Panel (a) shows the mean % DNA methylation status of hippocampal GR (Nr3c1) and MR (Nr3c2) genes. Panel (b) shows the mean % DNA methylation status of the NFGI-A binding site, a specific sequence present on the GR gene. All data refer to adult male mice perinatally exposed to nicotine (NIC, n = 5), cotinine (COT, n = 6) or just the vehicle (CTRL, n = 6). *p (adjusted for multiple comparisons) < 0.05; n.s., not significant.