Theta oscillation synchronize VTA and mPFC during ethanol-induced conditioned place preference and stress-evoked anxiety

The existence of addiction and anxiety comorbidity has been wildly known in clinic. Anxiety might stand as a potential risk factor for the addiction, as individuals may seek drugs or alcohol as a form of self-medication to mitigate their emotional symptoms [1, 2]. Conversely, the prolonged use of substances can intensify anxiety symptoms or even lead to the new anxiety disorders. These phenomena allow us to hypothesize whether there are similar neural mechanisms behind anxiety and addiction. Considering that both of the two psychiatric disorders exert significant influences on individual health and public well-being, understanding their intricate connections is crucial for the development of effective preventions and treatments.

Addiction and anxiety have their own complex neural mechanisms and are also interrelated. Previous research has shown that both addiction and anxiety are linked to the mesolimbic dopamine system, a crucial area for reward, motivation, addiction, and associated cognitive processes [3, 4]. The ventral tegmental area (VTA) dopamine neurons are the central hub of the mesolimbic dopamine system. Serving as the principal dopamine supplier, the VTA extends its dopaminergic fibers broadly to the other brain regions such as the nucleus accumbens (NAc), medial prefrontal cortex (mPFC), basolateral amygdala (BLA), and hippocampus. Among them, mPFC, a cortical region with various cell types, has reciprocal projection with VTA and plays a crucial role in the regulation of reward and emotion [5]. For example, injection of N-methyl-D-aspartate (NMDA) receptor blocker AP-5 into mPFC can enhance morphine-induced conditioned place preference (CPP) and significantly increase the activity of VTA dopaminergic neurons [6]. Furthermore, recent research has shown that activation of the circuit from prelimbic (PL), a subregion of the mPFC, to VTA exerts anxiolytic effects, indicating that the PL plays a distinct role in anxiety-like behavior [7].

Local field potential (LFP) signals in brain regions, recorded by in vivo electrophysiology system, reflecting regular and synchronized activities within the neuronal ensembles [8]. The LFP signals encompass diverse oscillatory patterns, which dynamically alter in response to varying conditions [9]. Network oscillations reveal efficient communication pathways among distributed neuronal populations [10], which is another synergistic mechanism between the two brain regions [11]. Previously published study has shown that theta-range (4–12 Hz) LFP signals in the prefrontal- amygdalar circuit are associated with anxiety [9]. Anxiety triggers an augmentation of theta-range LFP power within the mPFC-BLA-ventral hippocampal circuit [12,13,14]. Furthermore, simultaneous firing of VTA and mPFC neurons was observed when the mice received the sugar reward [15]. During the memory retrieval stage in cocaine-induced CPP, simultaneous LFP recordings from the mPFC, NAc, BLA, CA1 and VTA suggested the presence of short-term beta band (15–25 Hz) neural synchrony between these brain regions. This synchrony is modulated by the phase of slower (4 Hz) oscillations originating from the VTA [16]. These studies have demonstrated that the theta oscillations of mPFC and VTA may be involved in addiction and anxiety. Nonetheless, there is still a lack of systematic research comparing the similar and different manifestations of theta oscillations in the mPFC and VTA in these two psychiatric disorders. Addressing this issue is crucial for understanding the physiological mechanisms and developing the therapeutic strategies for them.

In this study, we simultaneously record the LFP signals from the mPFC and VTA in the ethanol-induced CPP and the three-chamber vicarious social defeat stress (VSDS)-evoked anxiety animal models. By analyzing LFP signals at different stages of behavioral patterns, we found a dominant theta band (4–12 Hz) oscillation in the mPFC and VTA during both the two models. Furthermore, analysis results of Granger causality showed that the coherent theta oscillations synchronize the VTA and mPFC in a bottom-up direction. Our experiments have illuminated the role of theta oscillations and their synchronized activity in the mPFC and VTA, highlighting their involvement in addiction and anxiety.

Animals

The 8-week-aged male C57BL/6 mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. and were randomly group-housed (four mice per cage). All mice were maintained under conditions of stable temperature (22–25 °C) and consistent humidity (50 ± 5%) with ad libitum access to food and water on a 12-h light/dark cycle (lights on at 8:00 AM). The sample size for each group was estimated based on previous experience and published studies. Age-matched mice in the same condition were randomly assigned to treatment groups and control groups. The sample sizes for each group are presented in the Table S1. All experimental procedures involving mice were approved by the Animal Care and Use Committee of Huazhong University of Science and Technology.

Electrode implantation

Prior to electrode implantation, mice were anesthetized using isoflurane. Anesthesia was induced with an inhalational dose of 3–4%, followed by a maintainable dose of 1–3%. Mice were secured in a stereotaxic frame, and their body temperature was maintained at 36.5 °C using a heating pad. The skull was leveled using bregma and lambda as landmarks, and craniotomies were performed using the anterior-posterior (AP), medial-lateral (ML), and dorsal-ventral (DV) coordinates derived from bregma: VTA (AP: −3.15 mm; ML: +0.6 mm; DV: −4.50 mm), mPFC (AP: +2.20 mm; ML: +0.3 mm; DV: −2.0 mm).

To target the mPFC and VTA, 16-channel array electrodes were implanted simultaneously. These electrodes consisted of 16 individually insulated nichrome wires (eight channels for each brain regions). The inner diameter of each wire was 35 μm, and their impedances ranged from 300–900 kilohms. The wires were arranged in a 2 × 4 configuration, with a channel spacing of 150 μm. The ground electrode was placed above the cerebellum, and the cerebellar drill hole was anchored with a stainless-steel screw and silver wire ground. The implanted electrodes were secured with dental cement. Following surgery, animals were singly housed under a 12-h light/dark cycle until their body weight returned to pre-surgical levels.

In vivo electrophysiological recording

Local field potential (LFP) signals from the VTA and mPFC were recorded using a multi-channel data acquisition system (Zeus, Bio-Signal Technologies, China) with a sampling frequency of 30 kHz and an LFP collection down-sampling frequency of 1000 Hz. One week after recovery from surgery, mice were habituated to the recording harness for 15 min before the LFP signals was recorded in homecage or behavioral tests, including OFT, EPMT, and ethanol-induced CPP. The recorded LFP data were analyzed using NeuroExplorer software, Chronux toolbox, BSMART toolbox, and MATLAB.

In vivo electrophysiological data analysis

The signal processing and analysis pipeline in this study was as follows: Raw neural signals were first band-separated using a fourth-order bidirectional Bessel filter (zero-phase distortion), with local field potentials (LFPs) subjected to low-pass filtering at 200 Hz and action potentials (spikes) subjected to high-pass filtering at 300 Hz to preserve the phase characteristics of neural oscillations. A narrowband notch filter with a bandwidth of 1 Hz was applied to selectively suppress 50 Hz power frequency interference. LFP power spectral density (PSD) analysis was performed in NeuroExplorer v5.424, with 1024 frequency points, using a Hanning window (50% overlap) to reduce spectral leakage, single tapering to optimize frequency resolution, and normalization expressed as a percentage of total power. Cross-regional coherence analysis was conducted by calculating coherence values, with parameter settings consistent with those used in PSD analysis, and coherence between the VTA and mPFC regions was smoothed in the time-frequency ___domain using a Gaussian filter (width = 3). Theta-band (4–12 Hz) Granger causality analysis was used to determine the directionality of influence between the two brain regions. According to pre-established criteria, the mouse data were excluded if the implantation site was incorrect through the anatomical validation. Based on signal quality, we excluded some channels and took the average across the remaining channels in each brain region. Cross-channel coherence and correlation analyses ensured consistency of the selected LFPs.

For spike signals, High-pass-filtered spike data were imported into Offline Sorter for spike detection via voltage threshold settings, waveform feature extraction, and clustering using the K-means algorithm to distinguish neuronal clusters across multiple channels. Cluster validity was assessed by checking waveform consistency and ISI histograms (to exclude refractory period violations). The collected neurons were classified based on the typical firing patterns of them [17, 18]. The spike signals of putative dopaminergic neurons in the VTA and putative pyramidal neurons in the mPFC were exported into NeuroExplorer for statistical analysis of spike width and frequency.

In vivo optogenetic manipulation

For in vivo optogenetic experiments, a mixture of AAV2/9-DIO-ChR2-mCherry and AAV2/9-TH-Cre viruses was injected into the VTA. A dual-electrode assembly integrated with an optical fiber was then implanted into the VTA and mPFC, with the fiber tip localized to the VTA. Following a 21-day viral expression period, ChR2 was activated using a 473 nm light (5 ms pulses at 8 Hz, 5 s on/5 s off). Concurrently, electrophysiological recordings were performed in both VTA and mPFC.

In vivo chemogenetic manipulation

In the hM4Di-mediated inhibition experiments, mice were divided into two groups: hM4Di+Saline and hM4Di+CNO. AAV2/9-DIO-hM4Di-mCherry virus was bilaterally injected into the VTA, while AAVretro-TH-Cre virus was bilaterally injected into the mPFC to achieve selective expression of hM4Di in the VTA neurons projecting to the mPFC. Electrodes were implanted in both the VTA and mPFC. After allowing for 14 days of virus expression, the mice were received the same CPP or VSDS training. Thirty minutes prior to the start of the test phase, mice in the hM4Di+CNO group were injected with CNO, while mice in the hM4Di+Saline group were injected with an equivalent volume of saline. During the experimental procedures involving the CPP, OFT and EPMT, LFP signals were simultaneously recorded from both the VTA and mPFC regions.

Immunofluorescence staining

Following anesthesia, mice were transcardially perfused with 4% paraformaldehyde. Brains were post-fixed overnight at 4 °C, dehydrated sequentially in 20 and 30% sucrose solutions, and sectioned coronally at 35 μm using the cryostat microtome. Sections were washed with PBST (0.05% Tween-20 in PBS), permeabilized with 0.3% Triton X-100, and blocked with 3% goat serum/0.1% BSA in PBS at 37 °C for 1 h. Primary antibody incubation with anti-tyrosine hydroxylase (TH; 1:1000, Proteintech, #25859-1-AP) was performed at 4 °C overnight. After three rounds of PBST washing, the brain slices were incubated with CoraLite 488-conjugated goat anti-rabbit secondary antibody (1:200, Proteintech, #SA00013-2) under light-protected conditions at 37 °C for 90 min. Nuclei were counterstained with DAPI before mounting with fluorescent mounting medium. Slides were stored at 4 °C until imaging analysis.

Ethanol-induced conditioned place preference (CPP)

The experiment of ethanol-induced CPP was conducted in a three-chamber place conditioning apparatus provided by Xinruan (Shanghai), featuring distinct tactile and visual stimuli. Each compartment of the apparatus had unique wall colors (black and white) and floor patterns (striped and circular). The training and test procedures are similar with previous reports [19, 20], which consisted of three phases: pre-training test (Day 1), training (Days 2–9), and post-training test (Day 11). During the pre-training test (Day 1), mice were placed in the apparatus to freely explore all the chambers for 20 min. The time spent in the black and white compartments was recorded to assess their natural preference for the two compartments. In the training phase (Days 2–9), mice were restricted to one compartment for 5 min after intraperitoneal injection of 2 g/kg ethanol (in a 20% saline solution) or an equivalent amount of saline. Training was conducted once daily, alternating between ethanol and saline. The control group received saline injections throughout. In the post-training test (Day 11), mice were allowed to freely explore all three compartments for 20 min. The apparatus was thoroughly cleaned with 70% ethanol after each test to eliminate odors and traces of previous mice. The time spent in the two compartments during the post-training preference test was recorded. The preference score was calculated as the percentage of the difference in time spent on the ethanol-paired side versus the saline-paired side to the total time. The mean time was determined by dividing the time spent on the conditioning side by the total time. On Day 11, LFP signals were recorded in synchrony during the post-training test, with the CPP apparatus enclosed by copper mesh to minimize electromagnetic interference.

Three-chamber vicarious social defeat stress model (VSDS)

Three-chamber VSDS procedures are similar with our previously published studies [21, 22]. The training process was conducted within a customized cage, measuring 545 mm in length, 395 mm in width, and 200 mm in height, which was divided into three chambers with perforated plexiglass partitions. An aggressive CD1 mouse was resident in the middle chamber, while two C57BL/6J male mice were placed in the other two chambers, one of which served as an experimental mouse (VSDS mouse) and the other as a demonstrator mouse. During the training phase, the demonstrator mouse located in the right chamber was introduced into the middle chamber and subjected to social defeat by the CD1 mice. Without receiving direct physical stimuli, the VSDS mouse in the left chamber underwent vicariously emotional stress by observing the demonstrator mouse that received repetitive social defeats in the middle chamber. After a 5-min social defect stress, the demonstrator mouse was placed back to the right chamber. Subsequently, the three mice stayed in their individual chambers, allowed for visual, olfactory, and auditory interactions, while physical contact was restricted, until the subsequent raining session commenced 24 h later. Whole training process was lasted for ten days. Control mice were housed with CD1 mice in the three-chamber cage, but received no physical or emotional stimuli.

Elevated plus maze test

The elevated plus maze test (EPMT), consisting of two open arms and two closed arms (length 350 cm each), was elevated to a height of 50 cm. Mice were allowed to acclimatize to the test room for at least 30 min before the experiment. Mice were connected to the electrophysiological recording equipment and placed in the center of the EPMT. The behavioral trajectories, LFP signals and the time spent in each arm during the test were recorded using a video tracking system over a 5-min period. The maze was cleaned with 70% ethanol to eliminate odors after each mouse. The proportion of time spent in the open arms were calculated as indicators of anxiety-like behavior. The EPMT was enclosed with copper mesh to shield against electromagnetic signals.

Open field test

During the open field test (OFT), each mouse was placed in an open field box measuring 50 cm in length and width and 38 cm in height, and observed for 5 min. The box was constructed from white high-density plastic. The center zone, defined as a square region occupying 36% of the total area (3 × 3), was calculated based on the entire testing area (5 × 5). Mice behavioral trajectories were recorded using a video tracking system (SuperMaze⁺, Xinruan, Shanghai). The time spent in the center zone was calculated as an indicator for assessing anxiety-like behavior. After mice were connected to the LFP collection equipment, they were placed in the open field arena. To reduce electromagnetic interference, the open field was surrounded by copper mesh.

Quantification and statistical analysis

The normality of all datasets was assessed using the Shapiro-Wilk test and homoscedasticity was checked using Bartlett’s test. Two-tailed Student’s t-test or non-parametric Mann-Whitney test was conducted for experiments with two groups. One-way ANOVA with post hoc Tukey’s test was performed for experiments with more than two groups and a single factor. Two-way ANOVA with post hoc Tukey’s test was performed for double-factor experiments.

All results were presented as mean ± SEM and were deemed statistically significant at *p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001. The P-values were calculated using GraphPad Prism 8 (Graph Pad Software, Inc.). All statistical details are provided in Data S1. The same investigators analyzed the data.

Theta power in the mPFC and VTA increases during exposure to ethanol-related environment cues

Previous studies have demonstrated that the theta oscillations (4–12 Hz) in several cortical brain regions implicated in cognitive functions such as reward processing and goal-directed behavior [23, 24]. We therefore investigated the role of theta oscillations in addiction and anxiety. In order to explore the neural dynamics in mPFC and VTA in addiction-related behavior, we conditioned mice using a classical ethanol-mediated CPP paradigm while permitting concurrent LFPs recordings of the two brain regions using in vivo electrophysiology (Fig. 1A). The microwire electrodes were implanted into the mPFC and the VTA. The locations of individual electrodes were confirmed by a postmortem histological analysis (Fig. 1B, Figure S1A, B). Following an appropriate recovery period, mice were subjected to behavioral experiment. The ethanol-induced CPP paradigm provides an indirect measure of drug reward in mice, which the training and test procedures are adapted from a previous report [19, 20]. Male C57BL/6 J mice were exposed to distinct compartments of CPP box marked by environmental cues, and given ethanol or saline in alternating sessions. After training, the cues trigger neural response without ethanol, which causes cravings. In the post-training test day, mice freely accessed to the whole box. The mean time spent in the conditioned side (paired with ethanol) and the preference score were utilized to evaluate alcohol-induced CPP. Consistent with previous study [20], the mice receiving ethanol injection (EtOH group) exhibited an increase of time spent on the ethanol-conditioned side compared with the control mice (Saline group) which solely received saline injections (Fig. 1C, D). There is no difference between the two groups on the distance travelled during the post-training test (Fig. 1E).

A Schematic illustration of the implantation locations of electrode and the timeline of the experimental protocol for CPP treatment in Saline group and ethanol (EtOH) group. CON, conditioned side (paired with ethanol); UN, unconditioned side (paired with saline). B Representative brain slices displaying electrode tracks in the VTA (Bottom) and the mPFC (Top). Scale bar, 500 μm. C Statistics of mean time spent in the conditioned side during pre-training test and post-training test. T_CON: time spent in the conditioned side; T_UN: time spent in the unconditioned side. D Statistics of preference scores during pre-training and post-training test. E Statistical analysis of the total distance traveled during post-training test. F Representative LFP traces and their filtered derivatives (4–12 Hz) from the mPFC and the VTA. G–J Statistics of mean theta Power spectral density in mice housed in homecage on day 10. G, H Power spectral density of mPFC and statistics of mean theta Power spectral density of mPFC; I, J The mean Power spectral density of VTA and histogram summarizes of relative power in VTA at theta band. K–N Statistical analysis of mean mPFC K, L and VTA M, N theta power for mice during CPP post-training test. Data are represented as the mean ± SEM. **p < 0.01, ***p < 0.001, and ns for no significance. Unpaired t test for C, D, H, J, L, N.

Full size image

On day 10, LFP signals were simultaneously recorded from the mPFC and the VTA using an electrode assembly in both homecage and the post-training test. The representative raw LFP signals and filtered 4–12 Hz rhythm in mPFC and VTA was shown as Fig. 1F. The power spectral density (PSD) analysis suggested that there was no significant difference of theta oscillation between Saline and EtOH mice in their homecage (Fig. 1G–J). In the CPP box during post-training test, EtOH group exhibited a prominent increase in the theta power compared to the Saline group (Fig. 1K–N). We also analyzed LFP signals in other frequency bands, such as the gamma oscillations in the 40–80 Hz range. The results showed no significant differences in gamma oscillations between the two groups, regardless of whether the recordings were taken in the homecage or the CPP box (Figure S2A, B). Additionally, we analyzed the neuronal spike signals that were simultaneously collected from the mPFC and VTA. The spike analysis revealed that the firing patterns of putative pyramidal neurons in mPFC and putative dopaminergic neurons in VTA were consistent with the LFP changes, further supporting our findings on the role of theta oscillations in ethanol-induced CPP (Figure S3). These findings demonstrate that power of theta oscillation is increased in both the mPFC and the VTA when mice that have experienced ethanol-induced CPP training are re-exposed to an environment related to ethanol.

Vicarious social defeat stress significantly elevated theta power in the mPFC and VTA

We next examined whether emotional stress could elicit the similar changes of LFP signals in the mPFC and VTA caused by ethanol-induced CPP. The three-chamber vicarious social defeat stress (VSDS) model was used [21]. Briefly, the experimental mouse was subjected to the daily emotional stress by observing the demonstrator mouse that received directly social defeat stress for continuously 10 days (Fig. 2A). Control mice were housed with CD1 mice in the three-chamber cage without physical or emotional stimuli. After 10 days training, the mice were subjected to the open field test (OFT) and elevated plus maze test (EPMT) to assess VSDS-induced anxiety-like behavior, accompanied by simultaneous electrophysiological recordings. The ___location of the electrode was validated by histological analysis (Fig. 2B, Figure S1C, D). In the OFT, the open field arena is divided into the center zone and the corner zone, while the EPMT apparatus has open arms and closed arms. The behavioral results showed that the time spent in center zone of OFT and the time spent in open arms of EPMT significant decreased in VSDS group compared with Control group (Fig. 2C and D). Meanwhile, there was no significant difference of the total distance in two groups (Fig. 2E and F). These results indicated that VSDS mice manifest the anxiety-like phenotype without affecting the locomotor activity. The electrophysiological results showed that mice subjected to VSDS showed elevated theta power in both the mPFC and the VTA at homecage (Fig. 2G–K), while no significant difference were showed in gamma power (Figure S2C). These findings suggest that theta power within the mPFC and VTA may be involved in VSDS-evoked anxiety-like behavior.

A Schematic illustration of the implantation locations of electrodes and experimental scheme of VSDS procedure and behavioral tests. B Electrode localization in the mPFC and VTA. The tip of the electrode is localized in the mPFC (Top) and VTA (Bottom). Scale bar, 500 μm. C Percentage of distance traveled in the center zone during the OFT. D Percentage of time spent in the open arms during EPMT. E Statistics of the total distance traveled during the OFT. F Statistics of the total distance traveled during the EPMT. G Raw (black) and theta-bandpass-filtered (green) LFP recording from the mPFC and VTA. H–K Statistics of theta power spectral density in mice housed in homecage on Day 11. H, I Power spectral density of LFP in the mPFC region and statistics of theta power. J, K The power spectral density of LFP in VTA and the statistical analysis of theta power. Data are represented as the mean ± SEM. **p < 0.01, ***p < 0.001, and ns for no significance. Unpaired t test for C–F, I, K.

Full size image

Enhanced theta power in the mPFC and VTA during cue-induced ethanol-seeking in both the conditioned side and unconditioned side of CPP box

While LFP power changes in EtOH group were observed in the CPP box after ethanol-mediated CPP training compared with the Saline group, there still remains a question that whether distinct oscillation patterns emerge in the mPFC and VTA when mice are situated in different compartments of the CPP apparatus. After ethanol-induced CPP training, the mice associated one side of the CPP box with ethanol which has rewarding effect to mice (ethanol-paired side), and the other side with saline (saline-paired side). Thus, we hypothesize that when mice were located exclusively in the ethanol-paired side, the mPFC and VTA may exhibit heightened theta power compared to the Saline group, whereas mice in the saline-paired side did not. We further analyzed the LFP signals from mPFC and VTA in mice when they were in ethanol-paired side and saline-paired side respectively during the post-training test. Contrary to our hypothesis, the theta power in the mPFC and VTA regions of the EtOH group mice was consistently higher than that of the Saline group, regardless of whether the mice located on the ethanol-paired side (Fig. 3A–H) or the saline-paired side (Fig. 3I–L). Collectively, our findings indicate that theta power in the mPFC and VTA increases when the mice were tested in the CPP box during the whole post-training test, rather than only the ethanol-paired side.

A–D Heatmap depicting the mean theta power of LFPs in mPFC A, B and VTA C, D in each position during the post-training CPP test. E–H Statistical analysis of mean theta power of mPFC E, F and VTA G, H in mice located in CON side between Saline and EtOH group. I–L Statistics of mean theta power spectral density in the mPFC I, J and VTA K, L when the mice were located in UN side. Data are represented as the mean ± SEM. ***p < 0.001, and ns for no significance. Unpaired t test for F. Mann-Whitney test for H, J, L.

Full size image

Heightened theta power of the mPFC and VTA in response to the anxiogenic environments

Given that our previous results found that mice subjected to the VSDS model exhibited increased theta power of the mPFC and VTA in homecage, we next investigated that whether the theta power changed in VSDS and Control mice during the exposure to anxiogenic environments. The theta power in the mPFC and VTA across the open field arena and elevated plus maze were analyzed. During exposure to the center zone of OFT, the LFP signals originating from the mPFC and VTA exhibited more robust theta oscillations in the VSDS group compared to the Control group (Fig. 4A–L). Similar results were observed in the EPMT, where mice demonstrated higher theta power in the mPFC and VTA while the mice were in the open arms (Fig. 4M–X). The spike signals were also consistent with the changes of LFPs (Figure S4). These results suggest that the theta power of the mPFC and VTA was obviously increased in response to the anxiogenic environments, including the center zone of OFT and the open arms of EPMT.

A–D Representative heatmap illustrating the real-time changes of theta power in mPFC A, B and VTA C, D across the open field arena. E–H Statistics of mean theta power spectral density of mPFC E, F and VTA G, H in mice located in center zone of open field arena. I–L Statistics of mean theta power spectral density of mPFC I, J and VTA K, L in mice located in corner zone. M–P Heatmap showing the real-time changes in theta power spectral density from LFP of mPFC M, N and VTA O, P across the elevated plus maze. Q–T The mean theta power spectral density of mPFC Q, R and VTA S, T located in open arms of EPMT. U–X Analysis of theta power spectra close arms in the mPFC U, V and VTA W, X during EPMT. Data are represented as the mean ± SEM. **p < 0.01, ***p < 0.001, and ns for no significance. Unpaired t test for F, H, R, V, X. Mann-Whitney test for J, L, T.

Full size image

Ethanol-induced CPP mice showed higher theta power in the mPFC and VTA before and after entering the conditioned side and un conditioned side

Increased theta power of mPFC and VTA in CPP box suggested their involvement in ethanol-seeking behavior. However, previous results show that EtOH mice showed higher theta power in both the ethanol-paired side and saline-paired side during the post-training test (Fig. 3). These results suggest that changes in mPFC and VTA oscillations are not simply explained by where the mice stayed in the CPP box. We further analyzed whether LFP patterns are associated with the mice entering into the conditioned side (CON, ethanol-paired side) and unconditioned side (UN, saline-paired). We extracted and analyzed 4–12 Hz LFP signals of CON and UN for 2 s before and after entering saline-paired side. The results showed that the EtOH group exhibited a notably higher theta power compared to the Saline group in three periods such as before CON (Fig. 5A–H), before and after UN (Fig. 5M–X) expect the phase after entering the CON, in which both the EtOH mice and the Saline mice displayed horizontal theta power (Fig. 5I–L). These findings indicate that mice undergoing ethanol-induced CPP showed an increased theta power in mPFC and VTA compared with Saline group before and after entering into CON and UN side.

A–D The theta power spectral density heatmap of mPFC A, B and VTA C, D when the mice enter into the CON side between the Saline and EtOH group. E–L Theta power spectral density and statistics of mean values in the mPFC and VTA before E–H and after I–L entering the CON side for 2 s in the Saline and EtOH mice. M–P The heatmap of theta power spectral density in the mPFC M, N and VTA O, P when the mice enter into the UN side for the Saline group G and EtOH group H. Q–X Comparison of theta power spectral density of mPFC and VTA before Q–T and after U–X the mice entering the UN side for 2 s. Data are represented as the mean ± SEM. *p < 0.05, ***p < 0.001, and ns for no significance. Unpaired t test for F, J, L, R, T, V. Mann-Whitney test for H, X.

Full size image

VSDS mice showed increased theta power in the mPFC and VTA when their companions were attacked

The previous results suggest that LFP signaling in mPFC and VTA is involved in anxiety-like behavior (Fig. 4). Therefore, we examined the LFP pattern in the single VSDS treatment. We found that the onset of anxiolytic events, when the VSDS mice observed their conspecific partners attacked by CD1 mice, significantly increased theta power of mPFC (Fig. 6A–C) and VTA (Fig. 6D–F) in the VSDS mice. These findings further reveal that the theta power in the mPFC and VTA was obviously elevated in VSDS mice experiencing anxiolytic events.

A–C Time-course changes of theta power spectral density from mPFC recordings in VSDS mice before and after the demonstrate mouse being attacked by the aggressor CD1 mice on Day 10 during the VSDS training. D–F Heatmap and statistical graphs displaying the power spectral density changes of theta power in the VTA of the VSDS mice before and after the CD1 aggressor attacks the demonstrator mouse. Data are represented as the mean ± SEM. *p < 0.05, ***p < 0.001, and ns for no significance. Unpaired t test for C, F.

Full size image

VTA drives mPFC in theta oscillation for both ethanol-induced CPP and VSDS-evoked anxiety

It has been found that mPFC is used for long range communication with its subcortical targets through the theta range oscillation [13]. Our results suggest that the similarity of the raw LFP traces from the mPFC and VTA, which implies the LFPs from these two areas might be highly coherent, in both the ethanol-induced CPP and VSDS-evoked anxiety animal models. We extracted the theta power (4–12 Hz) of mPFC and VTA in CPP, OFT, and EPMT at each time point to analyze the correlation between them. In the post-training test of ethanol-induced CPP, we found increased correlation in theta-frequency power between the mPFC and VTA in the EtOH group, compared to the Saline group (Fig. 7A–C). We also analyzed the mPFC and VTA coherence during post-training CPP test and observed a significant increase at 4–12 Hz before entering into CON (Fig. 7D–F). Moreover, the Granger causality index at the 4–12 Hz band in the direction from VTA to mPFC was significantly higher than that in the direction from mPFC to VTA (Fig. 7G, H).

A, B Representative plots of theta power correlation plots for VTA-mPFC from the 5-min recording data during the post-training CPP test in Saline group A and EtOH group B. Each data point represents the mean of theta power in the 1 s timewindow. C Statistical analysis of theta power correlations of mPFC-VTA between the Saline group and EtOH group. D–F Heatmap D and statistical analysis E, F showing the average values of the VTA-mPFC coherence before and after the mice enter the CON side during the post-training CPP test. G, H Granger causality analysis of the mutual influences (directionality) between the VTA LFP and the mPFC LFP prior to the mice entering the CON side. The blue and black curves denote the Granger causal influences from the mPFC to the VTA and from the VTA to the mPFC, respectively. I, J Representative examples of theta power correlation plot in homecage between VTA and mPFC from a 5 min recording for Control group I and VSDS group J after VSDS training. Each data point represents the mean of theta power during a 1 s window. K Changes of the averaged correlation r² of theta power in Control group and VSDS group. L, M Coherence analysis of VTA-mPFC in mice within the open and closed arms during the EPMT following VSDS training. N, O The results of Granger causality analysis illustrating the causal directions between the VTA LFP and mPFC LFP when the mice were located in the open and closed arms of EPMT. The green curve indicates the causality from the mPFC to VTA, and the black curve showing VTA to mPFC. Data are represented as the mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, and ns for no significance. Unpaired t test for C, F, K, M. Mann-Whitney test for H, O.

Full size image

Similarly, we also analyzed the raw LFP signals data to clarify the relationship of theta power between VTA and mPFC in anxiety-like behaviors. During the mice were located in the center zone of the open field arena, the VSDS group showed a higher correlation of theta power compared to Control mice (Fig. 7I–K). In the EPMT, the LFP coherence between VTA and mPFC also showed an increase in the open arms, compared to close arms (Fig. 7L, M). Meanwhile, the Granger causality analysis exhibited that the direction from VTA to mPFC was robustly higher than that in the direction from mPFC to VTA (Fig. 7N, O). These results suggest that the correlation of theta power between mPFC and VTA were increased in both ethanol-induced CPP and VSDS-evoked anxiety, and VTA drives mPFC in theta oscillation.

To further explore the causal relationship between VTA activity and the theta oscillations in the mPFC, we performed optogenetic activation of VTA dopaminergic neurons (VTA^DA) using 8 Hz blue-light (473 nm) stimulation. Previous studies have shown that 8 Hz optogenetic stimulation could pace theta oscillations [25]. VTA^DA neurons constitute the majority of neurons in the VTA and play a crucial role in both addiction and anxiety-like behaviors [20, 21]. The results revealed that selective activation of VTA^DA at 8 Hz significantly increased theta power in the mPFC (Figure S5), which supported our previous results that the VTA drives the mPFC in the theta band. Moreover, we discovered that chemogenetic inhibition of VTA^DA neurons projecting to the mPFC effectively reversed both ethanol-induced CPP and anxiety-like behaviors caused by VSDS. This intervention also simultaneously reduced the elevated theta oscillation power in the mPFC and VTA that we observed in these two behaviors (Figure S6, S7).

More and more clinical evidences demonstrate that some Mental Health Disorders (MHD), including anxiety, is always closely associated with Alcohol Use Disorder (AUD) [26]. Despite this well-documented association in clinic, relatively few studies have delved into the potential shared neural mechanisms underlying these two psychiatric conditions. Given that individuals with comorbid alcohol addiction and anxiety often require treatments that address both conditions simultaneously, understanding the common neural substrates of addiction and anxiety has the potential to significantly enhance therapeutic strategies and outcomes of the two disorders.

In this study, utilizing the ethanol-induced CPP and the 3C-VSDS-evoked anxiety animal models, we explored the neural correlates of addiction-related behavior and anxiety-like behavior. This was achieved by simultaneously LFP recording from the mPFC and VTA using in vivo multi-channel electrophysiological tool. Our findings indicate a substantial correlation of theta band oscillations (range from 4–12 Hz) between mPFC and VTA in ethanol-induced CPP and the 3C-VSDS-evoked anxiety mice, and this correlation was predominantly characterized by a directional influence from the VTA to the mPFC. These results provide novel insights into the theta oscillation deficits between VTA and mPFC in addiction and anxiety, as well as the potential shared neural mechanisms behind them.

Theta oscillations in addiction

The ethanol-induced CPP animal model was used to explore the addiction-related behaviors. CPP is a widely used model to assess the reinforcing effect of drug in rodents [27]. After repeated exposure with ethanol intraperitoneal injection, the contexts and cues associated with the ethanol administration can develop incentive motivational properties akin to ethanol abuse, triggering ethanol craving and relapse [28]. Previous studies have noted the effects of alcohol exposure on neural activity in several brain regions, such as hippocampus, cingulate cortex, and PFC [29, 30]. However, few comparative studies have directly explored the oscillation patterns in these regions underlying addiction-related behaviors. Our findings indicated that mice subjected to ethanol-induced CPP exhibit increased power in the 4–12 Hz band within the mPFC and VTA when re-exposed to an ethanol-associated environment.

It’s indeed intriguing to note that the theta power in the mPFC and VTA before the mice entering the ethanol-associated chamber was higher than after they had entered the chamber. This pattern of increased theta power was not observed in the saline-associated chamber. This finding was similar with earlier research, where a pre-entry increase in fluorescence calcium signals was also recorded, reflecting an anticipatory rise as the mice approached the ethanol-paired chamber [20, 31]. This could potentially indicate an anticipatory motivational response to ethanol-related cues in ethanol-induced CPP mice.

Theta oscillations in anxiety

In this study, we also investigated theta oscillation changes in the mPFC and VTA during anxiety-like behaviors. The three-chamber VSDS model was employed to elicit anxiety-like behaviors in mice, and this method has been validated in our previously published studies [21]. The cage, partitioned into three chambers by perforated plexiglass, housed an aggressive CD1 mouse alongside two C57BL/6J mice, with one serving as an observer and the other as an intruder subjected to daily attacks by the CD1. After 10 days of exposure to the emotional stress, the observer mice exhibited increased time spent in the corner zone of the OFT and decreased time in the open arms of the EPMT, both of which are indicative of anxiety-like behaviors.

Concurrent recording of LFP signals in mPFC and VTA during OFT and EPMT showed higher theta power and synchronous changes, when mice were in the center zone of the open filed arena, as well as the open arms of elevated plus maze. Previous studies have indeed demonstrated the role of theta oscillations in anxiety. Notably, research indicates that the distant communication between the mPFC and ventral hippocampus in the theta frequency is specifically enhanced under anxiogenic conditions. Also, the enhanced theta band activity significantly correlates with their behavioral performance in the OFT and EPMT [12]. Interestingly, mice in the Control group also displayed elevated theta power in the open arms of the EPMT (Fig. 4). This phenomenon can be attributed to the inherent nature of the open arms as an environment designed to provoke a certain level of anxiety in mice. However, this enhanced theta power was not as pronounced as in the OFT, which may be due to the open arms of elevated plus maze may induce a higher degree of anxiety compared to the center area of the open field arena.

In contrast to the findings from the CPP experiment, the VSDS mice also exhibited an increase in theta power within their homecage following a 10-day exposure period. The sustained theta power may suggest that mice were in a persistent state of anxiety. While our analysis of LFP signals revealed that the surge in theta power in the observer mice was transient and synchronized with the occurrence of attack event during the vicarious social defeat exposure. It is plausible that prolonged exposure to stress could lead to alterations in brain structure and function. These alterations might, in turn, perpetuate enhanced theta power, potentially contributing to a chronic state of anxiety.

Neural correlates of addiction and anxiety

Our findings suggest that the similar neural oscillation patterns were observed in addiction-related and anxiety-like behaviors. There is a strong correlation between the mPFC and VTA in the theta frequency range of 4–12 Hz among mice exposed to ethanol- or anxiety-related environments. Moreover, the Granger causality analysis indicates that the VTA clearly drives the mPFC in the 4–12 Hz band in both the two models.

In previous studies, there is a view that addictive drugs can trigger bottom-up, involuntary signals originating from the amygdala, leading individuals to make poor choices regardless of future adverse consequences, further resulting in the initiation and maintenance of drug addictive behavior [32]. The impulsive system, with the amygdala as a key region, can regulate or even hijack the ventromedial prefrontal cortex, a critical neural structure that plays a role in initiating the affective or emotional signals associated with long-term outcomes [32]. Meanwhile, other study has also shown that the ventral hippocampus neurons transmit/ anxiety-related information to the mPFC [33]. In anxiogenic environments, such as the open arms of elevated plus maze and the center zone of open field arena, the firing patterns of mPFC neurons are precisely entrained by the theta oscillations from the ventral hippocampus [12]. These findings suggest that in both addiction and anxiety-like behaviors, there is a phenomenon where lower-level brain regions (amygdala or hippocampus) modulate and even hijack higher-level brain regions (prefrontal cortex) in a bottom-up manner. These evidences provide some support for the conceptual premise that the neocortex brain layer continues to be dominated by the reptilian brain layer in the ‘Triune Brain’ theory, which is introduced by Paul MacLean in the 1960s [34].

This is consistent with our results, which have demonstrated that addiction or anxiety-related signals from the VTA are transmitted to the mPFC, a cortical area responsible for cognitive functions, attention and emotion, through synchronized theta oscillations. The theta rhythm affects the decision-making and cognitive functions, leading to the occurrence of addiction and anxiety. This could be a potential shared mechanism underlying these two psychiatric disorders.

More and more research has shown a close relationship between neuronal oscillations and psychiatric disorders. Under the abnormal behavioral states, the relevant brain regions may exhibit unique oscillation patterns and neural correlations, indicating that the neural oscillation patterns of the brain could potentially serve as a valuable tool for detecting and classifying mental illnesses. By recording the electroencephalographic activity of patients and combining it with other methods such as functional magnetic resonance imaging (fMRI) [35], we can conduct a more detailed diagnosis and classification of the psychiatric diseases, allowing for more precise treatment in clinic. Moreover, using exogenous rhythmic stimulation to intervene in the aberrant neural oscillations is also an effective method for treating mental disorders. Currently, several techniques are employed clinically, such as repetitive transcranial magnetic stimulation (rTMS), transcranial direct current stimulation (tDCS) [36], deep brain stimulation (DBS) and rhythmic sensory stimulation. Clinical study have indicated that applying 10 Hz rTMS to target the midline prefrontal cortex of individuals with methamphetamine use disorder can reduce mPFC beta activity, which is associated with reduced cravings [37]. Exogenous rhythmic stimulation is also widely used in the treatment of anxious individuals. In addition to traditional rTMS which has been proven to significantly alleviate patients’ anxiety and depressive symptoms [38], the theta burst stimulation (TBS), a new form of rTMS, has been found to improve anxiety and depressive symptoms by increasing the activation level of the prefrontal cortex [39, 40]. DBS in the bed nucleus of the stria terminalis (BST) significantly reduces anxiety symptoms in patients with treatment-resistant obsessive-compulsive disorder (OCD) [41]. These emerging applications of neuromodulation demonstrate the effectiveness of exogenous rhythmic stimulation in the treatment of mental illnesses. Compared to traditional drug therapy and psychotherapy, the use of exogenous rhythmic stimulation can selectively modulate specific neural oscillations, thereby more precisely affecting the brain regions and neural networks associated with specific mental illnesses, offering a more precise and efficient approach than traditional methods.

Certainly, there may be some possible limitations in this study. Although the ethanol-induced CPP model and VSDS-evoked anxiety model have been widely used in the previous studies into the mechanisms underlying addiction and anxiety, it is important to recognize that these models do not manifest the entire spectrum of addiction and anxiety. Mice treated with the different procedures may exhibit the similar behavioral performance, but the underlying neural mechanisms might be different in certain aspects. Thus, this may be an important issue for future research.

In conclusion, this study indicates that the theta oscillation of VTA and mPFC significantly increased in both the ethanol-induced CPP and VSDS-evoked anxiety mouse models, and the VTA clearly drives the mPFC in theta band. These findings point to an avenue of research toward a deeper understanding of how the theta oscillation in the VTA and mPFC is involved in psychiatric disorders of addiction and anxiety.