Depression and anxiety are the most common mental health problems, frequently appearing as comorbid illnesses (Pollack, 2005; Hranov, 2007; Campbell-Sills et al., 2012; Thibaut, 2017; Choi et al., 2020; Meuret et al., 2020). This phenomenon is of particular relevance since comorbid cases have a worse clinical outcome than each illness alone, with increased threat sensitivity and hyperarousal (Ironside et al., 2023), and a greater deterioration of the quality of life of individuals (Zhou et al., 2017; Saha et al., 2021). In addition, although the depressive disorder is the strongest predictor of suicidal behavior, comorbidity with anxiety increases the possibility markedly, up to 30%, of suicidal plans or attempts (Nock et al., 2010, 2013). However, it is still in discussion whether both disorders represent a common construct. Thus, research on the underlying altered brain networks and neurocomputation can help understanding their neurobiology and advance toward a personalized psychiatry (Chen, 2022).

One of the most consistent and remarkable findings in anxiety and depression is the existence of sex differences (Holden, 2005; Maeng and Milad, 2015; Jalnapurkar et al., 2018) in prevalence rates, symptomatology or disease course (Riecher-Rössler, 2017). The prevalence of depression and anxiety is consistently overrepresented in women than in men. According to the WHO, the prevalence of depression corresponds to 5.1% in females compared to 3.6% for males, and the prevalence of anxiety disorders is 4.6% in women compared to 2.6% in men (World Health Organization, 2019). Nevertheless, the most notable differences correspond to the comorbid cases of anxiety and depression, where women’s prevalence doubles that of men (McLean et al., 2011). In addition, women tend to have more symptoms in both disorders (Simonds and Whiffen, 2003; Schuch et al., 2014). Studies also report a more significant functional impairment and a higher probability of chronification and recurrence in both disorders in females (Marcus et al., 2008; Twenge et al., 2019; Höglund et al., 2020). Furthermore, some studies have shown that biological sex is an influential factor in antidepressant and anxiolytic pharmacological efficacy (Jalnapurkar et al., 2018; Eid et al., 2019). Also, in the last years, the study of sex differences in animal models has grown, leading a better understanding of the biological basis of mental disorders, but sometimes offering conflictive data for translating preclinical to clinical research (Tucker et al., 2022). However, sex as a biological variable remains underrepresented in this research field. This lack of research is more than surprising since the results of these findings could provide critical information for a better understanding of the etiological and pathogenic mechanisms of disorders which would probably lead to better therapeutic approaches (McCarthy et al., 2012). In this paper, we have used as a pharmacological anxiety model the drug FG-7142 (beta-carboline-3-carboxylic acid ethyl ester methyl amide). This compound has proven long-lasting anxiogenic capabilities (Skolnick et al., 1984; Thiébot et al., 1988) not only in animal models (Ninan et al., 1982; Ongini et al., 1983; Pellow and File, 1986; Lawther et al., 2015) but also in humans (Dorow, 1987; Horowski, 2020).

Despite the high relevance of depression and anxiety, current knowledge of the precise mechanisms underlying these pathologies are still poor compared to that of other chronic and debilitating diseases (Nestler and Hyman, 2010). At the cellular level, both pathologies share altered pathways, including mitochondrial impairment, which leads to reduced resilience under continuous stress (Tanaka et al., 2022). Between other alterations, depression and anxiety exhibit unbalanced neurotransmitters, impaired neuroplasticity, and abnormal neural circuitry (Pittenger and Duman, 2008; Liu et al., 2018). Identifying altered brain structures or circuits allows causal relationships with the disorders’ symptoms (Williams, 2016). In depression and anxiety, abnormal activity patterns arise in the amygdala-hippocampal-prefrontal network between other circuits. This network includes key structures in emotional and cognitive processing (McNaughton, 1997; Vanderlind et al., 2020).

Emotional processing largely depends on the activity of the amygdala and cortical areas located in the prefrontal regions (Etkin et al., 2015). Both anxiety and depression share a high negative hyperemotionality and an inability to regulate states and behaviors related to emotions and sensations (Elhamiasl et al., 2023). Neuroimaging studies have shown hyperactivation of the amygdala in response to threatening stimuli and negative emotional significance in depressed patients (Suslow et al., 2010; Victor et al., 2010). These abnormalities have also been observed in patients who, despite being under treatment, did not show any symptomatic remission (Drevets, 2001; Thomas et al., 2001; Suslow et al., 2010), in contrast with those who achieved an adequate response to the treatment (Hamilton et al., 2008).

In addition to emotional-related symptoms, depression and anxiety include cognitive alterations like memory dysfunction (primarily episodic and working memory), attentional biases, executive functions deficits, and cognitive and metacognitive vulnerabilities (Ferreri et al., 2011; Kircanski et al., 2012). While the management, integration, and recovery of memory depend on the hippocampus (Squire, 1992; Wallenstein et al., 1998), the executive, attentional functions and goal-directed behaviors have been related to the prefrontal cortex (PFC) (Koechlin et al., 1999; Miller et al., 2002). PFC dysfunction affects depressed patients’ memory and emotional learning capacity. Specifically, smaller volume and altered activity patterns of the ventromedial region of PFC (vmPFC) have been observed in patients with depressive disorder, contributing to fear extinction deficits (Battaglia et al., 2022).

The hippocampus contributes to emotional and cognitive processing, with a differentiated role along the longitudinal axis of the hippocampus (Strange et al., 2014). The dorsal hippocampus (HPCd) is key for learning and episodic memory (Moser et al., 1993; Moser and Moser, 1998) and spatial mapping (O’Keefe and Nadel, 1978). In contrast, the ventral hippocampus (HPCv) is involved in emotional memory, stress processing (Bannerman et al., 2004; Fanselow and Dong, 2010), and anxiety (Bannerman et al., 2004; Adhikari et al., 2010). However, its role is still unclear since discrete neuronal populations in the HPCv either promote or suppress anxiety through different projection targets (Parfitt et al., 2017) with sexual differences in mice (Wang et al., 2019). To date, little is known about the role of HPCi. However, this region has been proposed as a spatial, cognitive, and emotional integrator (Bast et al., 2009; Jin and Lee, 2021). Reduced hippocampal volumes have been observed in depression (Campbell and Macqueen, 2004; Videbech and Ravnkilde, 2004), correlating with greater severity and duration of the disease (McKinnon et al., 2009) and reversed by antidepressant drug treatment (Sheline et al., 2003). Abnormalities in the hippocampal microstructure also appear in comorbid anxiety with depression, associated with an exacerbated processing of threats (Cha et al., 2016). In turn, hippocampal damage or atrophy leads to poor stress processing, given its control over the hypothalamic-pituitary axis (Sapolsky, 2001).

Deep brain stimulation (DBS) consists of the implantation of electrodes in nuclei or bundles of brain fibers to supply currents that alter neural function therapeutically. Since Mayberg et al.’s (2005) seminal work, DBS has been applied in the anterior cingulate cortex (ACC) as a surgical approach in cases of treatment-resistant depression (Schlaepfer and Bewernick, 2013; Zhou et al., 2018; Crowell et al., 2019) and other targets like the median forebrain bundle (Rymaszewska et al., 2023). Recent meta-analyses corroborate its efficacy despite requiring more clinical and preclinical studies that facilitate a better understanding of its mechanisms of action (Khairuddin et al., 2020; Zhang et al., 2021). DBS has also been applied in the closest correlate in rodents to the ACC, the infralimbic cortex (DBS-IL), with antidepressant and anxiolytic effects in different behavioral tests (Gersner et al., 2010; Hamani et al., 2010; Hamani and Temel, 2012; Schmuckermair et al., 2013; Veerakumar et al., 2014; Jiménez-Sánchez et al., 2016; Torres-Sanchez et al., 2018).

Brain oscillations contribute to encoding, integration, retrieval, and transmission of neural information (Sejnowski and Paulsen, 2006). Therefore, aberrant brain oscillations would reflect altered neuronal dynamics or could lead to the development of a neurological or mental disorder (Başar, 2013), including depression and anxiety (Uhlhaas and Singer, 2012). In these pathologies, preclinical studies and clinical trials have focused mainly on the alpha, theta, and gamma waves (Knyazev et al., 2004, 2005; Smart et al., 2015; Fitzgerald and Watson, 2018; Fernández-Palleiro et al., 2020; Ippolito et al., 2022). There is also evidence of the alteration of slow waves, beta, and delta in depression and anxiety (Başar et al., 2001; Li et al., 2017; Muir et al., 2020).

We analyzed the neural oscillations of this network under the effects of the anxiogenic drug FG-7142 in male rats (Vila-Merkle et al., 2021). In that study, a characteristic communication pattern was observed. Slow waves disappeared, and a sustained theta emerged in hippocampal regions, with delta and beta oscillations in the HPCv and amygdala. This pattern was reversed by DBS of the infralimbic prefrontal cortex (DBS-IL).

However, this previous study was conducted solely in males. In the present paper, we compare the communication in the amygdalohippocampal network in male and female rats to contrast whether our prior observations are also valid in females. Most studies in animal models are conducted in male rodents, so research on females is also needed to understand how neural networks operate. We expected to find an increased response to the anxiogenic drug. In contrast to our expectations, when trying to reproduce these experiments in female rats, the results were remarkably different, with minor activity in the network. The present study also includes the analysis in both sexes of neuronal activation by c-Fos expression in several brain structures involved in stress, anxiety, depression, or fear in response to FG-7142, DBS-IL alone, and the combination of both. With this study, we aim to contribute to understanding the neurobiology of anxiety and depression and their comorbidity, highlighting the sexual differences in these animal models.

In this work, we compared the effects of FG-7142 administration in urethane-anesthetized male and female rats. The study complements prior observations only in male rats (Vila-Merkle et al., 2021). This drug has proven to have anxiogenic effects in humans and experimental models in various pharmacological, metabolic, neurochemical, and behavioral tests (Evans and Lowry, 2007). We found remarkable differences between both sexes, having much more intense effects in males.

In male rats, FG-7142 administration suppressed the characteristic slow wave pattern of the anesthesia and increased the peak frequency in all structures to the range of low theta or type 2 theta. This frequency increase was less marked in females. In this case, the peak frequency was around 1.2 Hz, in the delta range, and only significant in the dorsal and intermediate regions of the hippocampus.

Furthermore, only in males FG-7142 induced the appearance of information flows with high synchronization at low theta frequencies between the hippocampal regions. At the same time, the ventral amygdalohippocampal subcircuits exchanged information predominantly at coupled delta and beta frequencies. Females had a distinct communication pattern in the basal state which could prevent the alterations induced by the drug.

Finally, in both sexes, the drug increased c-Fos expression in brainstem structures related to arousal and stress like LC and DR, but NI activation only happened in males. In both sexes, at diencephalic levels, FG-714 hyperactivated the hypothalamic and thalamic paraventricular nuclei, highly related to behavioral control. Nevertheless, activation of the septal area, amygdala, hippocampus, and neocortical levels only occurred in the male group.

Our results in males, correlate well with the existing literature on the effects of the drug. FG-7142, in males, at similar and even lower doses than that used in this work induces pronounced anxiogenic effects on standardized tests such as the open field (Stephens et al., 1987; Dawson et al., 2006), the elevated plus maze (Pellow and File, 1986; Rodgers et al., 1995; Atack et al., 2005; Dawson et al., 2006), or the darkness-light test (Bueno et al., 2005). FG-7142 also alters the ability to regulate and inhibit emotional responses, potentiates fear behaviors, and induces avoidance deficits (Thiébot et al., 1991; Harris and Westbrook, 1998; Jelen et al., 2003; Kim and Richardson, 2007; Willadsen et al., 2018) and learned helplessness (Weiss et al., 1981; Drugan et al., 1985; Short and Maier, 1993; Maier and Watkins, 2005).

FG-7142 increases corticosterone levels in plasma (Pellow and File, 1985; Stephens et al., 1987) and activates brain structures involved in stress, including the amygdala, the HPCv, the bed nucleus of the stria terminalis, and the cingulate, prelimbic and IL cortices, and brainstem nuclei including the LC, the PAG, NI, or DR (Corda et al., 1983; Lyss et al., 1999; Singewald and Sharp, 2000; Singewald et al., 2003; Abrams et al., 2005; Funk et al., 2008). In turn, Hackler et al. (2007) observed an increase in the BOLD signal in the amygdala, the HPCd, and the hypothalamus in male rats for at least 40 min after administration of the drug. This latter effect matches the results of previous work in male rats that received FG-7142, in which the duration of the effect was more than 90 min without any spontaneous reversal observed (Vila-Merkle et al., 2021).

At the behavioral level, evidence supports the existence of sex differences in rodents in response to FG-7142. Early experiments (Meng and Drugan, 1993) showed that while male rats significantly decreased male open-field exploratory activity with FG-7142 at 5, 10, and 20 mg/kg, female rats needed a higher dose for lowering open-field activity (40 mg/kg). In contrast, Cottone et al. (2007) found that female rats with FG-7142 at the dose used in the present study (7.5 mg/kg) had anxiogenic-like behavior in the elevated plus-maze. Sex differences have also been described in other anxiety and depression models (Kokras and Dalla, 2014; Pavlova et al., 2020). In studies comparing female vs. male rats in the elevated plus maze, female rodents exhibit less anxiety-like behavior (Mora et al., 1996; Díaz-Véliz et al., 1997; Fernandes et al., 1999; Walf and Frye, 2007).

A major concern can be the possible influence of the hormonal stage in anxiety-like behavior. Previous findings suggest that proestrus or estrus may be anxiolytic (Mora et al., 1996; Díaz-Véliz et al., 1997; Fernandes et al., 1999; Walf and Frye, 2007). However, these concerns do not correspond to more recent observations made by Scholl et al. (2019). They aimed to clarify conflicting data regarding the anxiety-like behavior of female rats and the influence of the estrous cycle. Their study compared the behavior of male and estrous-staged female rats the elevated plus maze, open field, and a social interaction/avoidance paradigm. Regardless of the estrous cycle, female rats in the elevated plus maze spent more time in open arms, indicating lower anxiety. In the open field and the social anxiety test, however, they found no differences between males and females, and again, they found no influence of the estrous cycle.

In our experiments, the lack of control of the estrous phase could a limitation of our study to be considered in further studies. A detailed analysis of the effects of FG-7142 according to the estrous phase could clarify anxiety processing in female rats. However, not only female hormones influence anxiety-like behavior. Testosterone also contributes to sex differences in anxiety-like measures. Younger to middle-aged male rats, with higher testosterone levels, exhibit more anxiety-like behavior than females (Domonkos et al., 2018).

One of the most evident findings in this work is that during the anxiety-like states, slow-wave activity significantly decreased in all regions in males and only in the HPCd and HPCi in females. In both cases, acute application of DBS-IL restored the oscillatory pattern in every altered structure. Previous studies by our group have already demonstrated the effect of DBS-IL on slow waves. DBS-IL increased its power in HPCd and the BLA in anesthetized and in HPCv in FG-7142-treated male rats (Cervera-Ferri et al., 2016; Vila-Merkle et al., 2021). Etievant reported a similar effect in the infralimbic cortex of non-anesthetized rats after the DBS (Etiévant et al., 2015). In the present study, however, FG-7142 did not suppress the slow waves in the amygdala nor the HPCv in females. This different response can be related to a different cognitive response to stress paradigms. A distinct behavioral and cognitive response to stress has been observed in female rats, even with enhanced performance (McLaughlin et al., 2007; Bowman and Kelly, 2012).

Slow waves involve the synchronization of the membrane potentials of large distributed neuronal populations located mainly in the cerebral cortex and subcortical structures. These oscillations induce alternation between largely hyperpolarized states, characterized by neuronal silencing and depolarization periods (Van Someren et al., 2011). In mammals, including humans, the origin of slow waves seems to be cortical, since its generation depends on regions located in the PFC and anterior cingulate cortex, from where they are distributed to the rest of the brain in the anteroposterior direction (Massimini et al., 2004; Murphy et al., 2009). Slow waves are relevant to the maintenance of the general synchronization of the network, biochemical homeostasis of neurons, adaptation to environmental pressure, and memory consolidation (Maquet et al., 1997; Esser et al., 2007; Riedner et al., 2007; Neske, 2016).

The change from low-frequency oscillations to a sustained theta state suggests a general activation of the circuit in males and only in the dorsal and intermediate hippocampus in females. Theta activity is known for its involvement in the active states of the hippocampal formation, allowing memory processing (Hasselmo, 2005; Colgin, 2013) with dissociable oscillatory components for spatial and emotional cues (Wells et al., 2013). In rodents, theta has been subdivided into two bands: type 1 or non-cholinergic theta (referred to as high theta in this paper) and type 2 or cholinergic theta (referred to as low theta in this work).

Our results in males are consistent with previous studies finding theta type 1 rhythm mainly in the HPCd (Buzsáki and Moser, 2013), while type 2 theta activity depends mainly on HPCv (Gray, 1982; Kjelstrup et al., 2002; Pentkowski et al., 2006; Dringenberg et al., 2008). High theta or type 1 theta is related to sensorimotor integration (Bland and Oddie, 2001), locomotion (Vanderwolf, 1988; Fuhrmann et al., 2015), navigation, and spatial learning (Winson, 1978), while low or type 2 theta plays an important role in arousal, immobility and emotional behaviors, including defense, stress (Hegde et al., 2008; Jacinto et al., 2013, 2016), fear (Narayanan et al., 2007; Lesting et al., 2011; Seidenbecher and Lesting, 2012), and anxiety (Adhikari et al., 2010; Jacinto et al., 2013).

There is evidence supporting that theta activity in the HPCv is necessary for behavioral control during situations requiring high levels of arousal and anxiety in rodents, such as exploring unsafe places, the presence of predators, or decision-making in conflict situations (Bannerman et al., 2004; Adhikari et al., 2010; Seidenbecher and Lesting, 2012). Male rats familiar with an environment show decreased low theta rhythm in HPCd, HPCv, and BLA. However, anxious rats exposed to a novel environment exhibited increased theta power significantly in HPCv and BLA, and although to a lesser extent also in the HPCd (Jacinto et al., 2013), indicating that the increase in theta rhythm in these areas was related to the increase of anxiety levels. In contrast, reduced hippocampal theta activity is related to low-anxiety states and behavioral disinhibition while improving active avoidance behaviors in conflict-generating tasks (McNaughton et al., 2013). In addition, anxiolytic drugs reduce the frequency, power, and amplitude of the hippocampal theta rhythm, either under conditions of anesthesia or free movement (Zhu and McNaughton, 1994; McNaughton et al., 2007; Yeung et al., 2012).

FG-7142 induced an attentional or hyperarousal-like state, in which the activated amygdalo-hippocampal circuit appears to be processing the anxiogenic stimulus, with a very stable and sustained low-frequency theta. Although theta activity is necessary for the proper functioning of the hippocampus in memory and attention management, the dynamic changes of the oscillation are necessary for processing incoming information (Pomper and Ansorge, 2021). So, an excessively sustained theta could imply a block for processing other incoming stimuli. The cessation of theta activity is necessary to allow a change in focus of attention to switch attention (Kitchigina et al., 1999). A similar state arises following corticotropin-releasing hormone infusion into the dorsal raphe nucleus, which can be identified as a stress signal (Merino et al., 2021). This sustained activity could avoid the correct processing of new information and contribute to cognitive deficits associated with anxiety and comorbid major depression.

However, the different dorsoventral responses in females should be addressed in further studies to better delineate the behavioral correlate in stress management. The anxiogenic drug only increased theta activity in the HPCd. These results coincide with recent studies (Horváth et al., 2015; Ou et al., 2019) carried out in rats and mice of both sexes, observing a sexually differentiated response in the elevated plus maze. The authors found an increase in theta activity of HPCd in males and females, despite females showing lower anxiety-related symptoms in general. Their results would indicate that theta activity in the HPCd in females would not reflect a sustained state of anxiety without the presence of such activity in HPCv and BLA. This hypothesis is supported by the fact that optogenetic activation and inhibition of the BLA-HPCv projections increase and decrease anxiety-related behaviors in rodents (Felix-Ortiz et al., 2013).

In both sexes, the DBS-IL disrupted the persistent theta oscillations gradually causing their disappearance and returned the network to the slow-wave profile. It should be noted that depressed patients and animal models of depression exhibit a decrease in slow waves (Benca et al., 1992; Armitage, 1995; Dijk, 2010; Duncan and Zarate, 2013). Also, antidepressant therapies such as transcranial magnetic stimulation (Marshall and Binder, 2013; Binder et al., 2014) and antidepressant drugs not only increase the slow oscillations in the amygdalo-hippocampal-prefrontal circuit but also improve the management of cognitive and emotional information (Argyropoulos et al., 2009; Landsness et al., 2011; Duncan et al., 2013). Therefore, the action of the DBS-IL would suppress the pathological activity, reversing its effect by reinstating slow wave oscillatory activity, leaving the hippocampus and amygdala available for cognitive and emotional processing. However, since most works have been carried out in male rats, our work raises questions about the different oscillatory activity between both sexes.

Another relevant finding of this work was the sexually differentiated gamma activity in all experimental conditions. Females had much higher levels than males of basal gamma activity in all regions, coupled with slow waves and theta, and neither the administration of FG-7142 nor the DBS-IL induced a significant increase in gamma power in this group. In contrast, in male rats, with a lower basal gamma, the drug enhanced this activity in the more ventral regions, BLA and HPCv, and the DBS induced an even higher increase of gamma in all structures.

Gamma waves are fast-frequency oscillations which reflect the neuronal activity at the local level (Ray et al., 2008; Sirota et al., 2008). Clinical studies carried out with depressed and anxious patients (Lee et al., 2010; Fitzgerald and Watson, 2018) and animal models of depression (Sauer et al., 2015; Khalid et al., 2016) show a reduction in gamma in cortical areas, hence being a putative endophenotype of depression (Fitzgerald and Watson, 2018). Most studies on gamma activity in depression and anxiety focused on the PFC. However, a study combining postmortem tissue and imaging from living depressed patients saw a dysfunction of inhibitory interneurons, considered to be responsible for gamma oscillations (Devalle et al., 2017), not only in the PFC and the ACC but also in the amygdala and hippocampus (Fee et al., 2017).

In the hippocampus, gamma oscillations correlate closely with information encoding during memory formation and its coupling with theta oscillation is necessary for hippocampus-neocortex communication (Lisman and Buzsáki, 2008). In the amygdala, gamma oscillations are generated both during the environment evaluation for assigning emotional value and spontaneously in rumination processes (Headley et al., 2021). Thus, the increase in gamma activity with FG-7142 in HPCv and BLA could reflect an increase in information processing to assign an emotional value.

However, the further increase in gamma activity in the entire circuit during DBS-IL is challenging to interpret. However, it might be associated with the therapeutic effect since gamma power increases in numerous effective anxiolytic and antidepressant treatments. In the application of transcranial magnetic stimulation, antidepressant success correlates with enhanced gamma signaling during and after recovery (Pathak et al., 2016; Noda et al., 2017; Bailey et al., 2018).

A similar effect occurs with ketamine antidepressant levels (Hakami et al., 2009; Hong et al., 2010; Gilbert and Zarate, 2020; Zacharias et al., 2020). Ketamine at subanesthetic doses has a rapid antidepressant action like the DBS (Matveychuk et al., 2020) and is also effective for some anxiety disorders, particularly for comorbid anxiety and depression (Ionescu et al., 2014). The increase in gamma power after subanesthetic administration of ketamine correlates with the antidepressant responses, and the power of this frequency band increased markedly in responders (Cornwell et al., 2012). Also depressed patients with lower basal gamma showed better response to treatment and a noticeable increase in gamma power in multiple brain regions. In contrast, control subjects had baseline higher gamma activity and did not exhibit clear responses to treatment. In general, the authors suggested that an optimal amount of gamma power in the circuits involved in depression could be associated with euthymia, that is, without mood alteration (Nugent et al., 2019).

With this in mind, a high basal gamma activity in female rats could correlate with a homeostatic protective mechanism (Gilbert and Zarate, 2020) modulating the activation and inhibition of the circuit, in this case, avoiding the alteration of the activity of the hippocampus and the amygdala.

Again, we have observed sexually differentiated connectivity patterns. In females, we found no relevant changes in the coupling study in the experimental conditions. However, in males, FG-7142 decreased communication at slow wave frequencies throughout the circuit and increased phase coupling and communication at delta and beta ranges between the ventral regions and at theta band between most regions. We could hypothesize that under anxiety conditions, the coordinated activity of the ventral subnetwork (HPCi-HPCv-BLA) communicates via delta, low theta, and beta. At the same time, intrahippocampal communication operates in the low theta range. The DBS-IL also reversed these patterns by recovering the synchronized activity at a slow frequency.

Many studies have linked the theta coupling in the amygdalo-hippocampal-prefrontal circuit with anxiety states and innate fear in rodents (Adhikari et al., 2010; Lesting et al., 2011; Seidenbecher and Lesting, 2012; Stujenske et al., 2014). Direct manipulation of theta interactions within this circuit produces a freezing state in rodents suggesting a causal relationship between these oscillations and states of fear and anxiety (Courtin et al., 2014; Padilla-Coreano et al., 2016). Our results in males coincide with previous works correlating this phenomenon with the pathophysiology of anxiety. A recent human study has reported a similar increase in WPLI at theta frequencies in patients suffering from generalized anxiety (Xing et al., 2016). Also, in a study using an experimental hyperanxiety model, the increase in theta coherence between HPCv and BLA was considerably increased during entry avoidance behaviors in the open arms of the elevated plus maze than in the closed arms (Jacinto et al., 2013).

In males, DBS-IL increased synchronization between slow-gamma and theta-gamma in the circuit. We previously showed these effects between HPCd and BLA in anesthetized animals without FG-7142 (Cervera-Ferri et al., 2016). The increase of theta-gamma coupling in males after the DBS-IL also coincides with work carried out in patients with treatment-resistant depression who underwent DBS applied in the anterior cingulate cortex (Sun et al., 2015). Other clinical studies indicate that the increase in theta-gamma coupling in the neocortex and the hippocampus was related to improving the mnemonic processes related to working memory (Canolty and Knight, 2010). Likewise, subanesthetic administration of ketamine also increases theta-gamma coupling in the hippocampus (Korotkova et al., 2010; Caixeta et al., 2013). However, females did not show any change in these parameters. Further analysis of these measures in females during behavioral and cognitive tests in anxiety models would be of interest to better understand the existing differences.

We have also observed a sexually differentiated pattern in the study of neuronal activation by c-Fos expression induced by FG-7142, with females showing a was much more restricted activation at telencephalic levels and only sharing with males in the diencephalic and brainstem levels studied. As expected, many areas with higher c-Fos expression in response to FG-7142, especially the amygdalar complex, the HPCv, the PFC, the septal area and the paraventricular nuclei, are strongly involved in anxiety processing. It is well established that CeA and BLA are involved, in humans and experimental models, in the facilitation of fear and anxiety, respectively, and their inhibition significantly reduces these behaviors (Charney et al., 1998; Ventura-Silva et al., 2013; Ranjbar et al., 2017). Our study in males is consistent with c-Fos expression in male rats in response to different anxiogenic conditions, including FG-7142 (Singewald et al., 2003; Funk et al., 2008).

In our study, FG-7142 administration in males increased c-Fos expression in HPCv and HPCd CA1 fields but not in the other hippocampal subregions. These results agree with several studies, also in male rodents, in which exposure to anxiety-inducing environments and FG-7142 increased expression levels of c-Fos in these regions (Kurumaji et al., 2003; Singewald et al., 2003; Funk et al., 2008). In contrast, we found no significant changes in the female group in the hippocampal formation or the amygdalar complex. This lack of activation confirms the electrophysiological study and is consistent with a recent work showing no significant variations in c-Fos expression in the hippocampus and amygdala of female rats after exposure to the elevated plus maze or the open field, regardless of the estrous phase (Wang et al., 2019).

At neocortical levels, the anxiogenic drug induced elevated c-Fos expression in IL in both sexes. Furthermore, in males, we observed a generalized increase in prefrontal activity. Overall, the hyperactivation observed in males in this work coincides with a previous study that also demonstrated a generalized activation pattern in these structures in response to FG-7142 (Salchner et al., 2006).

FG-714 also induced an increase in c-Fos expression in the LSI, the MS and the VDB. Several studies on the functionality of the septal area document that the injury and inhibition of the lateral and medial septum are anxiolytic (Trent and Menard, 2010; Chee et al., 2015; Parfitt et al., 2017). Also, its activation increases anxiety-related behaviors (Huang et al., 2021). The activation of the MS/VDB complex in males can explain the neuronal activation in the hippocampus and the generation of hippocampal theta since this structure is considered its primary pacemaker (Gaztelu and Buño, 1982). In females, the lack of activation is also consistent with the electrophysiological results.

On the other hand, the anxiogenic drug increased in both sexes neuronal activity in the paraventricular nuclei, the DR and the LC. These structures promote the release of hormones and neurotransmitters involved in neuroendocrine and autonomic responses that mediate fear behaviors and psychological stress (Lowry et al., 2008; Ryan et al., 2011; Rajkumar et al., 2016; Kirouac, 2021; Zeng et al., 2021). In general, our results in the diencephalon and brainstem match previous works. However, the activation of NI exclusively in males is striking, without changes in females. Lawther et al. (2018) documented its activation after the administration of FG-7142, but that work was done exclusively in males. The NI constitutes the primary input toward the MS/VDB complex (Teruel-Martí et al., 2008), which in turn projects toward the hippocampus. Thus, the activation of the NI exclusively in males could justify the different activation at the septohippocampal level and the lower presence of theta rhythm in females. Also, the NI projects directly to the amygdala and ventral hippocampus (Goto et al., 2001; Olucha-Bordonau et al., 2003; Smith et al., 2010), which could justify a more substantial influence on these regions than even on the HPCd.

Very similar increases had been described in male rats exposed to anxiogenic environmental situations (Singewald and Sharp, 2000; Singewald et al., 2003). Our results also agree with other works correlating anxiety responses in the elevated plus maze, conditioned fear or aversive conditioning with increased expression of c-Fos in the amygdalar complex (Saldívar-González et al., 2003; Sorregotti et al., 2018), the hippocampus (Galvis-Alonso et al., 2010), the prelimbic and infralimbic cortices (Duncan et al., 1996; Berg et al., 2019), the septal area (Reis et al., 2011; Chee et al., 2015), the diencephalon and brainstem (Singewald and Sharp, 2000; Hale et al., 2008; Rajkumar et al., 2016; Kirouac, 2021; Toropova et al., 2021; Zeng et al., 2021).

In our work, the lower neuronal activation in females, especially in the amygdala, hippocampus, and septal area, could be related to less expression of anxious behavior in response to FG-7142 (Meng and Drugan, 1993). Our results of neuronal activation in response to the combination of FG-7142 and DBS-IL indicate that its ability to change the activation pattern induced by the anxiogenic state could mediate part of its therapeutic action. In our case, 1 h of DBS-IL alone in both sexes already increased local activation in some regions. Other works have documented similar local increases in c-Fos expression after stimulation of the IL (Veerakumar et al., 2014) suggesting that high-frequency stimulation causes specific activation in the area surrounding the stimulation electrode. In addition, DBS-IL alone increased the activation in other cortical areas, most notably in males. A recent study found that DBS-IL increased the expression levels of zif268, another early expression gene used to study neuronal activation, in the prelimbic, cingulate, ectorrinal, orbital and temporal (Bregman et al., 2015). Since all these structures maintain direct or indirect connections with IL, our results support that the DBS not only acts locally. Instead, the high-frequency stimulation can also modulate the neuronal activity of structures connected to the DBS target region. However, the activation observed following the DBS-IL alone was always lower than with the anxiogenic drug.

Regarding the changes in the levels of neuronal activation induced by the combined application of DBS-IL and FG-7142, our results showed that acute stimulation counteracted the activation of all those regions activated by the anxiogenic drug. These data are consistent with the findings of studies that examined c-Fos expression in animals exposed to unpredictable chronic stress situations. While exposure to stressful stimuli increased c-Fos expression of DR-mediated serotonergic pathways, high-frequency stimulation significantly reduced that activation (Lim et al., 2015). It would indicate that the therapeutic action of DBS-IL applied in a pathological state can modulate the aberrant activity of those structures that connect with the stimulated region.

FG-7142, a proven anxiogenic drug, originates a sexually differentiated response in the amygdalo-hippocampal network, which is key in emotional and cognitive processing. Surprisingly, the data point to a reduced response in female rats, contrasting with the higher prevalence of anxiety in women than in men. Also, our results show that DBS-IL can reverse aberrant oscillatory activity. Our study contributes to a better understanding of anxiety processing by the hippocampal formation and the amygdala, with a detailed study of the changes in the dorsoventral axis of the hippocampus, including the intermediate and ventral region, much less studied than the dorsal hippocampus. Also, our results point to the need to be careful in animal models’ translational studies of anxiety. Their different participation in pathological states and their therapeutic reversal offer new perspectives for future study.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

All the experimental protocols were followed according to the Animal Care Guidelines of the European Communities Council Directive (2010/63/E.U.) and approved by the Research Ethics and Animal Welfare Committee of the University of Valencia (A20200227130134, A20191121151842, and A1537174325669) and the Valencian Government (Generalitat Valenciana) before performing the experiments.

AC-F: conceptualization, project administration, supervision, and writing—original draft. HV-M and AG-M: data curation. HV-M: formal analysis. AC-F and AL: funding acquisition. HV-M, AG-M, RC-J, and AC-F: investigation. AB-S and AC-F: methodology. JM-R, AL, and AC-F: resources. VT-M, HV-M, AG-M, and AC-F: software. HV-M and AC-F: visualization. HV-M, AG-M, RC-J, JM-R, VT-M, AB-S, and AL: writing—review and editing. All authors have read and agreed to the published version of the manuscript.