Tetrodotoxin-mediated inactivation of medial prefrontal cortex decreases wakefulness and rapid eye movement sleep, and increases slow-wave sleep in rat

There is a large body of literature to support the subcortical regulation of arousal states [1], but relatively few studies have investigated the role of the cortex. Previously, we have demonstrated that cholinergic stimulation of the prelimbic region in the medial prefrontal cortex (mPFC) produced active emergence from general anesthesia [2] and increased wakefulness in unanesthetized rats [3]. We further demonstrated that the inactivation of rat mPFC facilitated anesthetic induction and delayed emergence from anesthesia, indicative of decreased arousal levels [4]. Based on these studies showing an arousal-promoting role of mPFC, we hypothesized that the inactivation of the prelimbic region in mPFC will decrease wakefulness and increase sleep. Therefore, in the current study, we determined the effect of pharmacological inactivation—via local tetrodotoxin (TTX) infusion—of mPFC on sleep-wake states.

The detailed methodological approach is provided in Supplementary section. In brief, adult Sprague Dawley rats (n = 7 male, 300–350 g, Charles River Inc.) were prepared for recording bilateral electroencephalogram (EEG) and electromyogram (EMG), and for bilateral microinjection of 0.9% saline (Saline—vehicle control) or 156 µM tetrodotoxin (TTX) into the mPFC (Bregma: anterior 3.0 mm, lateral 0.5 mm, ventral 3.0 mm) [5]. After at least 10–14 days of postsurgical recovery and habituation to the recording set-up, the rats were connected to the EEG/EMG recording cable between 4:00 pm and 5:00 pm and baseline recordings were conducted without any experimental intervention across 12-h dark and 12-h light period starting at 8:00 pm. On a separate day, the rats were similarly connected to the recording set-up and received 0.5 µL of either saline or TTX into the mPFC 30 minutes before the start of lights-off period (8:00 pm). The postmicroinjection recordings were conducted for 24 h starting at 8:00 pm. The saline and TTX sessions were conducted at least 5–7 days apart in a counter-balanced manner such that 3 out of 7 rats received saline infusion in the first session and the remaining 4 rats received TTX infusion in the first session. The sample size for the study and TTX concentration were based on previous studies from our [4, 6] and other laboratories [7, 8]. The EEG (0.1–300 Hz, 1-kHz sampling rate) and EMG (1–100 Hz, 1-kHz sampling rate) data were collected using a Grass model 15 LT bipolar 15A54 Quad amplifier (Natus Neurology) paired with an MP150 data acquisition unit (Acqknowledge 4.1.1, Biopac Systems Inc.). After the completion of recording sessions, the microinjection sites were verified histologically (Supplementary Figure S1). The data files were deidentified and manually scored (SleepSign, Kissei Comtec) in 4-second epochs into wakefulness (Wake: low-amplitude fast-wave EEG, high muscle tone), slow-wave sleep (SWS: high-amplitude slow-wave EEG, low muscle tone), and rapid eye movement (REM) sleep (low-amplitude fast-wave EEG, muscle atonia) (Supplementary Figure S2). The 4-second epochs were averaged in 3-h bins across the 24-h recording period. The EEG power spectral changes were analyzed and shown in Supplementary Figure S3 to confirm that the inactivation of mPFC did not distort the EEG signal quality and hence did not bias the sleep-wake state identification. Linear mixed models along with Tukey post hoc corrections for multiple comparisons were used to compare the effect of baseline, saline injection, or TTX injection, on the time spent in each state, mean duration/episode for each state, mean number of episodes for each state, EEG spectral power in each state, and the latency to the onset of SWS and REM sleep.

TTX-mediated inactivation of the mPFC produced a delayed decrease in the time spent in wakefulness [for 3–6 h: Baseline 79.10 ± 3.82% vs TTX 54.68 ± 10.21%, p = .004, t(138) = 3.299; for 9–12 h: Baseline 83.45 ± 4.91% vs TTX 47.64 ± 7.21%, p ≤ .0001, t(138) = 4.484, Saline 74.28 ± 5.74% vs TTX 47.64 ± 7.21%, p = .003, t(138) = 3.336; and for 12–15h: Saline 45.24 ± 6.05% vs TTX 27.58 ± 5.60%, p = .04, t(138) = 2.509; for 15–18h: Baseline 55.75 ± 4.90% vs TTX 36.07 ± 5.99%, p = .04, t(138) = 2.497] (Figure 1A). Simultaneously, there was a significant increase in the time spent in SWS [for 3–6 h: Baseline 17.16 ± 2.89% vs TTX 43.02 ± 9.10%, p = .02, t(145) = −2.853; for 9–12 h: Baseline 15.55 ± 4.50% vs TTX 52.03 ± 7.77%, p ≤ .0001, t(145) = −4.511, Saline 21.61 ± 4.62% vs TTX 52.03 ± 7.77%, p = .0008, t(145) = −3.722; for 12–15h: Saline 50.36 ± 6.63% vs TTX 74.66 ± 5.83%, p = .02, t(138) = −2.495; and for 15–18 h: Baseline 38.85 ± 4.29% vs TTX 63.57 ± 6.94%, p = .007, t(145) = −3.079, Saline 39.96 ± 5.15% vs TTX 63.57 ± 6.94%, p = .007, t(138) = −2.767] (Figure 1B). TTX injection into mPFC produced a significant and sustained decrease in REM sleep in 0–3 h bin [Baseline 10.05 ± 1.47% vs TTX 3.25 ± 1.15%, p = .001, t(154) = 2.964, and Saline 10.92 ± 2.06% vs TTX 3.25 ± 1.15%, p = .0006, t(153) = 3.790], 6–9 h bin [Saline 9.78 ± 1.14% vs TTX 3.08 ± 1.50%, p = .004, t(153) = 3.279], 12–15 h bin [Baseline 5.40 ± 1.05% vs TTX 2.05 ± 0.99%, p = .03, t(154) = 2.593], 18–21 h bin [Baseline 10.33 ± 1.10% vs TTX 4.85 ± 1.54%, p = .02, t(154) = 2.761, and Saline 11.10 ± 2.05% vs TTX 4.85 ± 1.54%, p = .01, t(153) = 2.613], and 21–24 h bin [Baseline 8.29 ± 0.89% vs TTX 3.98 ± 1.05%, p = .04, t(154) = 2.402, and Saline 11.08 ± 1.04% vs TTX 3.98 ± 1.05%, p = .0006, t(153) = 3.798] (Figure 1C). Most of the time, the decrease in the time spent in wake state and the increase in the time spent in SWS appear to be driven by nonstatistical changes in the mean duration and/or the mean number of episodes (Figure 1D, E, G, and H). The decrease in REM sleep was primarily due to a sustained decrease in the mean duration of REM sleep episodes (Figure 1F) [for 0–3 h: Baseline 64.86 ± 7.95s vs TTX 39.50 ± 8.53s, p = .04, t(153) = 2.369; for 6–9 h: Baseline 58.38 ± 9.19s vs TTX 23.63 ± 9.49s, p = .04, t(153) = 2.370, and Saline 46.75 ± 5.57s vs TTX 23.63 ± 9.49s, p = .02, t(153) = 2.752; for 9–12 h: Saline 54.63 ± 5.85s vs TTX 24.00 ± 6.48s, p = .02, t(153) = 2.752; for 12–15 h: Baseline 70.43 ± 9.17s vs TTX 11.13 ± 4.27s, p = .0001, t(153) = 5.724, Saline 37.88 ± 11.75s vs TTX 11.13 ± 4.27s, p = .02, t(153) = 2.742, Baseline 70.43 ± 9.17s vs Saline 37.88 ± 11.75s, p = .007, t(153) = 3.081; for 15–18 h: Baseline 61.71 ± 9.87s vs TTX 18.25 ± 7.59s, p = .04, t(153) = 2.429, Baseline 61.71 ± 9.87s vs Saline 35.75 ± 8.80s, p = .04, t(153) = 2.429; for 18–21 h: Baseline 67.86 ± 6.67s vs TTX 30.75 ± 10.78s, p = .002, t(153) = 3.531; and for 21–24 h: Baseline 60.43 ± 8.40s vs TTX 27.25 ± 6.47s, p = .006, t(153) = 3.142]. There was also a decrease in the mean number of REM sleep episodes (Figure 1I) in 0–3 h bin [Saline 30.75 ± 5.53 vs TTX 9.88 ± 1.92, p = .0002, t(153) = 4.126; Baseline 16.43 ± 1.07 vs Saline 30.75 ± 5.53, p = .02, t(154) = 2.794], and 21–24 h bin [Saline 32.13 ± 3.19 vs TTX 15.00 ± 3.40, p = .003, t(153) = 3.385; Baseline 15.14 ± 1.18 vs Saline 32.13 ± 3.19, p = .003, t(154) = 3.302]; there was a significant increase in the mean number of REM sleep episodes in 9–12 h bin [Baseline 1.95 ± 3.96 vs TTX 20.12 ± 3.71, p = .002, t(154) = −3.465] (Figure 1I). Despite the changes in the time spent in SWS and REM sleep, there was no significant change in the latency to the onset of either SWS or REM sleep (Supplementary Figure S4).

Figure 1.

Tetrodotoxin (156 µM) microinjection into the prelimbic region of the medial prefrontal cortex caused an immediate and lasting decrease in REM sleep, a delayed decrease in wakefulness (Wake) and increase in slow-wave sleep. (A–C) The percentage of time spent in wakefulness, slow-wave sleep, and REM sleep, during the 24-h baseline recording (triangles), and after 0.9% saline (circles) and 156-µM TTX (squares) microinjections. The mean duration (seconds) per episode is shown for wake (D), slow-wave sleep (E), and REM sleep (F). The number of episodes during the baseline and posttreatment period is shown for wake (G), slow-wave sleep (H), and REM sleep (I). The deep blue bar on the timeline (x-axis) indicates dark phase and the yellow bar indicates light phase of the 24-h recording period. The significance symbols denote p < .05. The exact p-values are provided in the text in the results section. ^#significant compared to baseline, *significant compared to saline injection. ^†significant difference between baseline and saline injection.

The most salient finding of the current study is the immediate and sustained decrease in REM sleep after pharmacological inactivation of the mPFC. We also report a decrease in wakefulness and an increase in SWS, both of which were primarily observed 9 h after the TTX injection, which could be due to the known increased drive for wakefulness during the lights-off period that could potentially counter the effect of TTX in depressing wakefulness. Such long-lasting changes after TTX-mediated inactivation of other brain regions have been reported previously [7, 8]. Our data align with previous reports from our [2–4] and other [9] laboratories that showed an arousal-promoting effect of the mPFC. Notably, our data complement and support the findings of a recent study in which optogenetic stimulation of pyramidal neurons in the infralimbic region of mouse mPFC was shown to increase REM sleep and wakefulness and decrease SWS, whereas optogenetic inactivation of mPFC pyramidal neurons decreased REM sleep [10]. In contrast, a previously published study showed that the inactivation of layer 5 neurons across the mouse cortex increased wakefulness and decreased SWS without any effect on REM sleep [11], which points to the selectivity of mPFC pyramidal neurons in modulating REM sleep. Another recent study showed that somatostatin-positive GABAergic neurons in mouse PFC project to the lateral hypothalamic area where the stimulation of the terminals of these neurons produced sleep-preparatory behavior and SWS [12]. An increase in SWS along with suppression of REM sleep after the activation of inhibitory GABAergic neurons in the mouse mPFC was also reported by Hong and colleagues [10]. Altogether, these studies demonstrate distinct roles for different neuronal populations within the mPFC in the regulation of sleep-wake states. In the current study, pharmacological inactivation of the prelimbic region within the mPFC decreased both wakefulness and REM sleep, which could be due to the diverse connectivity patterns of mPFC neurons with subcortical areas associated with sleep-wake regulation [13]. For example, Hong and colleagues [10] demonstrated that the mPFC pyramidal neurons projecting to the lateral hypothalamus regulate REM sleep, whereas another recent study [14] showed that the mPFC neurons projecting to the ventral tegmental area modulate wakefulness or behavioral arousal from anesthesia. Further studies are needed to map out the behavior-specific anatomical connectivity of mPFC neurons.

Our study is limited by a small sample size and the lack of data from female rats. In addition, we did not directly measure the inactivation of mPFC neurons, and the low temporal resolution associated with pharmacological inactivation precludes any inferences about state transitions. Despite these limitations and the limited scope of our study, our data extend the recent findings from a mouse study of infralimbic mPFC [10] to prelimbic mPFC in rats and contribute to the understanding of the prefrontal cortex as a key node in arousal circuitry.

Supplementary material

Supplementary material is available at SLEEP online.

Funding

This work was supported by the National Institutes of Health (Bethesda, MD, USA) grant no. R01 GM111293 to DP, and funding from the Department of Anesthesiology, University of Michigan Medical School, Ann Arbor. We thank Dr Chris Andrews (Consulting for Statistics, Computing & Analytics Research, University of Michigan, Ann Arbor, MI) for help with statistical analysis.

Disclosure Statement

Nonfinancial Disclosure: None.

Data Availability

Data will be made available on request.