5-MeO-DMT causes sleep-like brain wave patterns in memory and thinking areas of awake rats

A total of 17 adult male rats were used (Rattus norvegicus, Wistar, 250–350 g, ~ 2 months old). The animals were housed in an appropriate vivarium under controlled temperature and humidity, with lights on at 6:00 and lights off at 18:00. All methods were performed in accordance with the ARRIVAL guidelines and relevant regulations, including approval by the Ethics Committee at UFRN (permit CEUA #11/2015) and by the Brazilian Health Regulatory Agency (ANVISA AEP # 018/2021).

The study used 5-MeO-DMT (Merck) dissolved in DMSO (0.27mg/ul), and then diluted in 0.9% saline to final doses.

Electrodes were designed and handmade as described elsewhere^44,45. The electrode arrays were designed to be implanted in the mPFC (16 channels) and the hippocampus (16 channels) according to Paxinos Atlas⁴⁶. In each area, two rows of 8 electrodes were implanted along the anteroposterior axis, with 250 µm of inter-electrode space.

Experimental animals were subjected to surgery for electrode implantation as described in Souza et al.⁴⁵. Briefly, anaesthesia was induced with inhaled isoflurane at 5% followed by intraperitoneal ketamine and xylazine at respective doses of 100 mg/kg and 8 mg/kg. When needed supplemental doses of ketamine and xylazine were applied to maintain anaesthesia. Electrodes and the cannula were then implanted aiming at the dorsal hippocampus (Anteroposterior (AP): − 3.28 mm; Mediolateral (ML): 2.0 mm; Dorsoventral (DV): 2.2 mm); the prefrontal cortex (AP: + 3.72 mm; ML: − 0.5 mm; and DV: 3.5 mm); and the third ventricle (AP: − 0.96 mm; ML: 1.8 mm; and DV: 3.2 mm). Following the surgery animals were treated with analgesic, anti-inflammatory, and antibiotic drugs and given time to recover.

Animals were submitted to four, weekly-spaced experimental sessions in random order (Fig. 1A; 1 to 8 sessions per rat, totalling 54 drug and 20 saline sessions). Each experimental session consisted of three consecutive days that started with a one-hour recording of the animal in the arena (55.5 × 65 × 45 cm). On the second day, after the baseline recording, animals were injected with one of the three 5-MeO-DMT doses (for ICV groups: 100 ug, 150 ug, or 200 ug diluted in 2 ul of saline; for IP groups: 1 mg, 3 mg, or 10 mg diluted in 0.3 ml of saline) or saline (negative controls, referred to as 0 g ICV or IP) and recorded for an additional two to three hours. Electrophysiology and animal behaviour were recorded at each session. All experiments reported here refer to the second day of each experimental week. The experimental unit was the data obtained from each animal in each experimental session; Supp Table S1 specifies the number of experimental units allocated to each group. We selected a moderate sample size because the dose-dependent effects of 5-MeO-DMT in the HP and mPFC of freely behaving animals were evaluated for the first time in the present study, and therefore, the initial intention was to gather basic evidence regarding the dose-dependent effects of 5-MeO-DMT in more complex experimental designs. The allocation of experimental units to control and treatment groups was randomized using a computer based random order generator. ACS was aware of the group allocation during the allocation, the conduct of the experiment, and the outcome assessment; BCS was aware of group allocation during the data analysis. A few experimental units had problems with either the video of the behavior or the recorded LFPs. In those cases, the unit was analyzed only when appropriate. Experimental units in which the ICV injection showed any signs of leaking or injection resistance (due to cannula clogging) were completely excluded from all analyses. A few experimental units had problems with either the video of the behavior or the recorded LFPs. In those specific cases, the unit was only used in analysis that did not require the missing data (e.g., behavioral analysis when the electrophysiological recordings were missing). Moreover, experimental sessions in which the animal mostly slept after saline injection were excluded from the speed-controlled analyses due to lack of moving data points. For the slow and mid gamma analyses, only recording units that had at least one electrode containing the respective gamma were included. Identification of the slow or mid gamma channels was done by blind inspection of the average spectral decomposition triggered by theta peaks. Those exclusion criteria were defined a priori.

Before the first experimental session, animals were handled (~ 20 min) and habituated to the experimental room and box for at least three days (1 h per day). Experimental room temperature of 24 °C was constant for all sessions. Experiments were done in the afternoon or beginning of the evening (14–18 h) under dim light. For intracerebroventricular dosing, we used a Hamilton syringe of 10 ml connected to a polyethylene tube and a thinner cannula to fit in the implanted cannula. The volume dosing of 1ul unilaterally was manually administered for a duration of 30 s approximately.

The electrophysiological and video recordings were conducted using the Plexon Omniplex System or Intan Technologies coupled to a Logitech C920 webcam (30 Hz). Plexon and Intan signals were recorded at 40 kHz and 30 kHz, respectively, and down sampled to 1 kHz for LFPs. Intan recordings were synchronised to video using a TTL pulse in a pseudorandom frequency to the Intan and a LED was placed in the filming area of the camera. No video alignment pre-processing was done for Plexon data since it is automatically synchronised.

The animals were tracked using Cineplex from Plexon or IdTracker to detect the animals’ centre of mass. Once the tracking was performed, the animal’s speed was calculated. Further behavioural analyses were done using Cineplex Editor.

All the LFP analyses described below were performed in Matlab. Behavioural and electrophysiological changes during sleep were measured using ANOVA test followed by a Tukey–Kramer post-hoc test when applicable. Normality was tested using a Kolmogorov–Smirnov test. For speed (and power) controlled analyses custom subsampling procedures and generalized linear models (GLM) were used. The level of significance was set either at p < 0.05, p < 0.01 or p < 0.001.

We analysed the time-varying frequency content of the LFP signal using a spectrogram, to calculate the following outcome parameters: delta (0.5–4 Hz), theta (5–12 Hz), low gamma (25–50 Hz), mid gamma (50–100 Hz), spectral state maps, mean duration of each wake-sleep state, and transition probabilities across states. The signal was divided into 1-s overlapping segments, with a 50% overlap. The power spectral density (PSD) for each segment was calculated using Welch’s method, applying a Hamming window. The analysis used a sampling frequency of 1000 Hz and 4096 FFT points.

To address any potential behavioural differences between conditions, we employed a bootstrap procedure⁴⁷ to match the speed distributions of animals between conditions (baseline and post-injection sessions). Specifically, we randomly sampled half of the windows in a 30 min period post-injection and paired them with baseline windows based on the speed of the animals, thereby ensuring equivalent speed distributions across sessions. This procedure was repeated 500 times for each recording day. We then compared the mean delta and theta power of baseline and post-injection (e.g. Fig. 2C, F, I, L) conditions.

First, we quantified the behaviours after 5-MeO-DMT ICV dosing (Fig. 1B,C; Supp Fig. S1). Besides some stereotyped behaviours already mentioned in the literature (e.g., wetdog shake), we quantified abnormal behaviours through empirical observation. More specifically, we show here the 10 behaviours most observed across animals: intermittent and uncoordinated gaiting or jumping; backwards gaiting; flat body gaiting; turning on its axis; quiet; uncoordinated gaiting; still; jumps; head tremor; and wetdog shake (Supp Fig. S1 and Supp Table S2 for a more detailed description).

For each of those behaviours we measured the average time spent and the number of events (Supp Fig. S1) during 5-MeO-DMT or saline ICV injection. We found that 5-MeO-DMT ICV sessions (all doses together) led to a significant increase of mean time spent and the number of events of the behaviour ‘intermittent uncoordinated gaiting or jumping’, ‘uncoordinated gaiting’ and ‘still’ (Supp Fig. S1). To better understand and visualise the results we separated the behaviours in two groups: ‘still’ and ‘altered behaviours’ (a combination of behaviours 1 to 6), and evaluated the contribution of each dose to those two classes of behaviours (Fig. 1C, ANOVA with Tukey–Kramer post hoc test). We found that 5-MeO-DMT vs saline is significantly different for both behaviours (ANOVA, #p < 0.05, ‘Still’: Mean time: F(3,20) = 7.6747, Number of events: F(3,20) = 5.3164. ‘Altered behaviours’: Mean time spent: F(3,20) = 3.2362, Number of events: F(3,20) = 3.8690). More specifically, both the mean time spent and the number of events of ‘still’ are significantly decreased for every dose compared to saline. Conversely, there was an increase of the mean time and number of events of ‘altered behaviours’ for the dose ICV 200 ug compared to saline (Fig. 1C, Tukey–Kramer post hoc test, *p < 0.05; **p < 0.01).

We next evaluated the impact of 5-MeO-DMT on the spectral profile of hippocampal LFPs. As the drug significantly affected locomotion (Fig. 1B) and speed is known to influence LFPs, we employed a generalised linear model (GLM) to assess the effect of 5-MeO-DMT administration on delta (1–4 Hz) and theta (5–12 Hz) power while accounting for speed (Fig. 2, see Methods for details). To achieve this, we first used the GLM to predict the relative delta or theta power (compared to the baseline session) while controlling for speed. Our analysis showed that there was a significant and dose-dependent impact of 5-MeO-DMT on delta power in the hippocampus following IP injections (Fig. 2C–E, 5-MeO-DMT vs. saline: p = 0.001; Dose: p < 0.001). Interestingly, we did not observe the same effect for ICV injections. In contrast, we observed a significant decrease of theta power following the administration of 5-MeO-DMT using both IP (Fig. 2I–K) and ICV (Fig. 2L–N) delivery, with a significant dose-dependent effect specifically for ICV injections (5-MeO-DMT vs. saline: p < 0.001; Dose: p = 0.001). In summary, our findings indicate that 5-MeO-DMT administration leads to a decrease in theta power in the hippocampus for equivalent locomotion speeds, as well as an increase in delta power specifically following IP injections.

Hippocampal theta oscillations are known to nest different gamma oscillations that reflect distinct circuits and pathways (for a review see³⁶. In alignment with previous studies^47,49–53, we divided the gamma range in two components: slow (20–50 Hz) and mid (50–100 Hz). To assess the impact of 5-MeO-DMT administration on these two gamma rhythms, we conducted a matching analysis that controlled for changes in speed and theta power (Fig. 3A,B; see Methods). This analysis revealed a significant dose–response relationship between 5-MeO-DMT ICV injection and slow gamma power (Fig. 3C, 5-MeO-DMT vs. saline: p = 0.008; Dose: p = 0.013). Specifically, slow gamma amplitude tended to be lower when nested in theta cycles during the post-injection epoch, compared to the baseline with matched power and speed. We observed a similar dose–response effect for mid gamma power (Fig. 3D, 5-MeO-DMT vs. saline: p = 0.053; Dose: p = 0.027). Notably, our results indicate that IP injections had no discernible effect on either slow or mid gamma power.

We proceeded to investigate the impact of 5-MeO-DMT on gamma amplitude modulation by the theta rhythm in the HP. To address the decrease in theta power caused by the drug we integrated theta amplitude into our matching analysis, enabling us to examine any additional effects on modulation beyond this reduction alone. Specifically, we selected individual theta cycles from pre- and post-injection epochs with matching amplitude and speed. We then calculated the average of the envelope of each gamma band, aligned to those theta peaks (see Fig. 4A,C). To evaluate theta modulation, we computed the power spectra of the triggered average envelopes (see methods) and quantified the power within the theta range. A higher concentration of power within the theta range of the triggered envelopes indicates a stronger modulation between the theta phase and the gamma amplitude.

Following this analysis, we found that 5-MeO-DMT administration resulted in a significant reduction in the theta modulation of slow gamma oscillations specifically after ICV administration (5-MeO-DMT vs. saline: p = 0.002; Dose: p = 0.003). Furthermore, we found no effect on the theta modulation of mid gamma oscillations in the HP.

Next, we investigated the impact of 5-MeO-DMT on prefrontal cortex delta and theta power (Fig. 5) using the same analysis framework for the HP channels (Fig. 2). Our findings indicate that IP injections of 5-MeO-DMT led to a dose-dependent increase in delta (5-MeO-DMT vs. saline: p < 0.001; Dose: p < 0.001) and theta power (5-MeO-DMT vs. saline: p = 0.002; Dose: p < 0.001), while ICV administration resulted in a significant dose-dependent decrease of theta power (5-MeO-DMT vs. saline: p < 0.001; Dose: p < 0.001). These opposite effects suggest that the route of administration plays a critical role in the impact of 5-MeO-DMT on prefrontal cortex delta power. Further, we observed no significant effect for the theta range with IP injection, whereas ICV administration led to a dose-dependent decrease of theta power, consistent with the HP results.

To analyse the spectral patterns of 1-s windows of combined hippocampal and cortical activity, we used a manifold approach called state map³⁸. This approach generates a two-dimensional representation of LFPs by calculating the ratio of two frequency bands: (1) delta over theta plus delta, and (2) from delta to beta over delta to gamma (see Methods for details). The state map method was primarily designed to identify three main clusters of the sleep–wake cycle: WK, SWS, and REM. In summary, high amplitude theta power, i.e., low ratio (1), distinguishes WK and REM from SWS, whereas relative gamma power on theta oscillations separates WK from REM (WK theta cycles exhibit higher gamma power compared to REM cycles, captured by ratio (2).

We first used the state maps of the baseline session of each animal to define the three main sleep-wake clusters (see Supp Fig. S3,S4 for examples). Then, we projected the activity of post-injection windows onto that map (Fig. 6A and Supp Fig. S2A). Consistent with our previous results, this analysis revealed that animals exhibited an altered spectral profile after 5-MeO-DMT dosing. Once we had the state map for pre- and post-dosing periods for saline and drug experiments, we calculated the transition between each pair of states (Figs. 6B). We found that saline and 5-MeO-DMT groups had similar transition probabilities (Figs. 6C for ICV, Supp Fig. S2B for IP). Notwithstanding, transitions between WK-like and REM-like states tended to increase for 5-MeO-DMT ICV experiments—especially for the doses 100 ug and 150 ug (Fig. 6B; see the increase in the mean of transition probability for the WK purple bars in Fig. 6C).

Next, we computed the duration and animal speed (locomotion) for each sleep or wake episode. We found a significant increase in the time spent in WK state after 5-MeO-DMT ICV dosing, while it did not change for REM-like or SWS-like states compared to saline (Fig. 6D, left panel). Conversely, we found that mean speed across episodes was increased during REM-like and SWS-like states, while it decreased during WK after 5-MeO-DMT IP (Fig. 6D, right panel). Animals did not spend more time in a sleep state; they rather presented a greater speed during that sleep-like state. In that sense, we could say that there was a significant change in the state map profile that was not accompanied by the behavioural state. In other words, 5-MeO-DMT administration pushed the hippocampal activity closer to the states observed during baseline sleep, as indicated by post-5-MeO-DMT periods invading the SWS and REM clusters in the state map. Note in the representative examples the abnormal transitions (Fig. 6B, left panel) and high speed cluster displacement from WK, an effect not seen for saline (Fig. 6D right).

We found similar state map changes for the 5-MeO-DMT IP dosing, even though the transition probability between sleep-like and wake-like clusters did not change significantly (Supp Fig.), Likewise in ICV experiments, in the IP experiments there was an increase in time spent in WK and a decrease in the WK mean speed across episodes (Supp Fig.). The opposite occurred for the SWS-like state, for animals spent more time and at higher speed within the SWS-like state after 5-MeO-DMT IP dosing (Supp Fig.). In other words, the rats presented a SWS-like state even though they were actively moving across the open field. A similar yet non-significant trend occurred during the REM-like state; the REM-like regions at the dose of 10 mg had higher speed compared to saline (Supp Fig., right panel). In this case, animals not just presented a displacement of higher speeds cluster, but also such displacement occurred during a longer period and it seems to be towards the SWS-like cluster (or even to REM-like). Note that effect in the examples shown in Supp Fig.. Rats 3–5 presented a high-speed cluster closer to SWS-like compared to Rat5 during saline (Supp Fig., right-bottom panel). S2B S2C S2C S2C S2A S2A

Animals treated with 5-MeO-DMT exhibited a variety of altered behaviours (Fig. 1, Supp Fig S1), some of which are well documented stereotyped behaviours, while others were not previously reported in the literature. This may be attributed to the fact that most studies assess behavioural effects of drugs on specific tasks, rather than on spontaneous activity. Our findings indicate that the most significant changes in behaviour were related to ‘uncoordinated gaiting’ and ‘stillness’.

The ‘Uncoordinated gaiting’ behaviour observed in our study may be related to the ‘forepaw treating’ associated with serotonergic syndrome⁵⁴. Spanos et al. showed a dose-dependent increase in “forepaw treating” and ‘low body posture’, but not in ‘head weaving’ after MDMA injection in rats⁵⁵. Conversely, we did not observe a significant change in the behaviour equivalent to ‘flat body gaiting’ or ‘low body posture’.

In accordance with the appearance of altered behaviours, we found that animals exhibited a lower level of stillness (both in duration and frequency) following administration of 5-MeO-DMT at all doses compared to when they were given saline (Fig. 1B and C).

Although the remaining quantified behaviours did not change individually, they did show a general increase in prevalence following administration of 5-MeO-DMT. When grouping behaviours 1 to 6 as “altered behaviours,” we observed a significant increase in both the duration and frequency of these behaviours (Fig. 1C). This effect appeared to be dose-dependent, with a stronger impact observed at higher doses of the drug. While the behaviours of ‘backward gaiting’, ‘head tremor’ and ‘wetdog shake’ did not show significant changes when considered individually, we observed an increase in their prevalence when analysed together, consistent with observations reported in mice treated with d-LSD¹³. Wetdog shake, also known as head-twitch response (HTR), is a rapid and sudden side-to-side shaking of the head/neck that is a well-known marker of the agonistic effect of the serotonergic system^56,57. This behaviour is highly rhythmic and produces a wave-like oscillation around 90 Hz in mice⁵⁷. In this study, we observed the occurrence of ‘wetdog shake’ in rats treated with 5-MeO-DMT (Supp Fig. S1B), which is consistent with evidence indicating that HTR is increased by other psychedelics in mice, such as DOI, DOM, and N,N-DMT⁵⁸. However, we did not observe a significant change in the flat body gait behaviour (Supp Fig. S1), as previously reported for d-LSD¹³. Likewise, we found no significant increase in the duration of ‘turning around its axis’, a behaviour that has not been well described in the current literature on psychedelics, although 5-MeO-DMT tended to increase the number of events.

The differences observed between ICV and IP dosing may be related to the availability of 5-HT agonists, peripheral action, and the amount of drug that reaches the central nervous system. Our IP findings contrast with a similar study by Halberstadt et al.¹², which found no initial increase in locomotion, but reported a later hyperactivity⁵⁹. However, it is important to note that lower doses were used in their study, potentially leading to a subliminal or delayed effect on locomotion. The increase in locomotion we observed in the present study could be an effect of higher doses and more direct route of administration in the ICV experiments.

We found a significant decrease of hippocampal theta power in a dose–response manner that could not be explained by changes in animal speed. This is in accordance with the idea that the median raphe acts as a theta desynchronization nucleus^60–62. Moreover, the power of theta-associated gamma oscillations and their coupling to theta phase were significantly reduced by 5-MeO-DMT. The administration of 5-MeO-DMT induced significant alterations in animal locomotion, which is an important factor to consider when interpreting the potential effects observed in electrophysiology, as speed can have a profound impact. It is worth noting that our study differs from previous investigations in that we took this confounding variable into consideration.

The differences between IP and ICV dosing are also noteworthy. In the hippocampus, the dose-dependent increase in delta power following IP injections was not observed for ICV injections, implying a possible peripheral mechanism. Classical psychedelics such as 5-MeO-DMT can also bind to serotonergic receptors of the peripheral nervous system (PNS) and this could affect the CNS quite differently from the ICV dosing. Serotonin is a major controller of the gut motility, including intrinsic reflex, epithelial secretions, and vasodilatation. Furthermore, its signalling takes part in vagal extrinsic and spinal afferent fibre activation, leading to pancreatic secretion, satiation, pain and discomfort, nausea, and vomiting⁶³. The high availability of serotonin in the PNS provided by the IP dosing may have led to a peripheral increase in receptor activation, causing physiological alterations such as gastric discomfort. Although not life-threatening, those alterations could potentially change the subjective experience of the animal during the experience, as well as its neural correlates. A second non-exclusive hypothesis is that the peripheral binding of psychedelics causes fewer molecules to reach the CNS. Yet, these putative lower doses were still effective, as shown by the significant electrophysiological changes observed in the HP and mPFC for theta and gamma oscillations after IP injections of 5-MeO-DMT. Regarding theta-nested gamma oscillations, IP injections produced less dose-dependency than ICV injections, which suggests that IP injections either dampen or shift the dynamic range of the effective intracerebral doses within the HP or mPFC.

State maps based on LFP power ratios provide a consistent measure of state-dependent spectral features as animals alternate between sleep and waking³⁸. Under drug-free conditions, state maps usually display very similar patterns across animals, despite inter-individual differences in behaviour and LFP signals. In the present study the state maps presented marked alterations after psychedelic dosing compared to saline, with some changes shared by the two groups investigated (Fig. 6 and Supp Fig. S2). More specifically, ICV 5-MeO-DMT dosing did not influence the duration of SWS-like episodes but rather increased the animals’ speed during this spectral state (Fig. 6C). This result indicates that the animals treated with ICV 5-MeO-DMT were behaviourally awake, but their HP seems to be in an SWS-like state. There was also an increase in the duration of WK episodes, although the animals’ speed decreased after ICV 5-MeO-DMT (Fig. 6C). Therefore, when animals had electrophysiological features of WK, they were more likely to be at a lower speed after ICV 5-MeO-DMT compared to saline.

The results obtained in the IP 5-MeO-DMT experiments were overall like those from the ICV 5-MeO-DMT experiments, with animals spending more time in the SWS-like and WK-like spectral regions (Supp Fig., left panel), and presenting higher speeds during the SWS-like state and lower speeds during the WK-like state, when compared to animals treated with saline (Supp Fig., right panel). S2C S2C

One limitation of our study is that we conducted all experiments exclusively on male rats, aligning with the predominant use of male rodents in both behavioral (e.g.,^12–15,17) and electrophysiological (e.g.,^19,23) psychedelic research. Our primary aim was to confirm and build upon existing electrophysiological findings, leading us to maintain methodological consistency with previous research. Nonetheless, we recognize that this standard approach stems from the outdated belief that hormonal cycles in female rodents contribute to increased experimental variability, a notion recent literature has debunked^64–66. Given this, we acknowledge that the exclusive use of male rats is a limitation of our study, and future research should consider including female subjects to address this gap and enhance generalizability.

Another important limitation is that we did not investigate the receptors involved in the effects described. 5‐MeO‐DMT is mainly a nonselective serotonergic agonist, with some binding to the serotonergic and noradrenergic transporters, as well as dopamine receptors^59,67. Further studies are therefore needed to specify the receptor dependence underlying the effect.

The data will be made available by Sidarta Ribeiro upon reasonable request.