Adolescence is a sensitive developmental period during which the brain undergoes substantial maturation (Larsen and Luna, 2018) and is especially vulnerable to different insults including alcohol. Alcohol use is often initiated in adolescence (Morean et al., 2018), with over 2 million adolescents in the United States between 12 and 17 years of age initiating alcohol in 2019 (Substance Abuse and Mental Health Services -SAMHSA, 2020). During late adolescence alcohol use becomes heavier, often in the forms of binge (SAMHSA, 2020) and high intensity drinking (Patrick and Terry-McElrath, 2019). These more risky drinking patterns are associated with short-term cognitive deficits (Jones et al., 2018), however less is known about the long-term alterations as a result of adolescent alcohol exposure.

Preclinical rodent models have identified several long-lasting behavioral impairments associated with adolescent alcohol (ethanol) exposure (Crews et al., 2019), including, but not limited to, increases in anxiety-like behavior (Kyzar et al., 2019; Sakharkar et al., 2019; Varlinskaya et al., 2020), reduced learning, memory, and cognitive flexibility (Gass et al., 2014; Fernandez et al., 2017; Fernandez and Savage, 2017; Rico-Barrio et al., 2019; Macht et al., 2020; Varlinskaya et al., 2020), and also increased risk-taking (Nasrallah et al., 2011; Clark et al., 2012; Kruse et al., 2017). These behavioral impairments are also associated with numerous neurobiological changes such as decreased acetylcholine markers (Vetreno and Crews, 2018; Vetreno et al., 2020; Kipp et al., 2021), increased neural inflammation (Vetreno and Crews, 2012; Vetreno et al., 2017; Vetreno and Crews, 2018; Macht et al., 2020; Vetreno et al., 2020), and dysregulated epigenetic regulators (Kokare et al., 2017; Kyzar et al., 2017; Kyzar et al., 2019).

Our lab has found that adolescent intermittent ethanol (AIE) induces alterations of social behavior in adulthood (Varlinskaya et al., 2014; Varlinskaya et al., 2017; Dannenhoffer et al., 2018; Varlinskaya et al., 2020). These AIE-induced social impairments are sex-specific and evident only in males. These social deficits are reversible through systemic administration of either an oxytocin (OXT) agonist or vasopressin receptor 1b (AVPR1b) antagonist (Dannenhoffer et al., 2018). Furthermore, AIE leads to decreased hypothalamic OXT receptor and increased AVPR1b protein membrane surface expression (Dannenhoffer et al., 2018), specifically in males, findings likely contributing to the social deficits evident in these animals. However, our understanding of the neural mechanisms underlying AIE-induced social deficits is still limited.

Social behavior is regulated by various brain regions that together form the social brain network (Newman, 1999). The social brain network was originally outlined as being composed of the extended amygdala, lateral septum (LS), hypothalamus (HYPO), preoptic area, and midbrain (Newman, 1999). Although evidence supports the involvement of these regions in the regulation of social behavior, more recent evidence also indicates regions outside of this network that contribute to social behavior (Yizhar, 2012; Allsop et al., 2014; Vanderschuren et al., 2016; Ko, 2017). For example, social interaction (approach and contact with a social partner) in males elicits greater firing of neurons in the prelimbic (PrL) cortex (Minami et al., 2017), with similar findings of increased firing of projections from the orbitofrontal cortex (OFC) to basolateral amygdala (BLA) during social approach in males (Li et al., 2021). Furthermore, chemogenetic activation of PrL to BLA circuitry reduced social preference in females (Huang et al., 2020) whereas optogenetic activation of the PrL to nucleus accumbens (NAc) pathway in both sexes leads to decreased social investigation (Murugan et al., 2017). Lesioning the PrL, thus disrupting all afferent and efferent pathways of the PrL, leads to increased social interaction among males (Shah and Treit, 2003) although baclofen and muscimol mediated inhibition of the PFC, including the PrL, infralimbic (IL), and OFC, reduced social play behavior in adolescent males (Van Kerkhof et al., 2013). Taken together, recent evidence suggests that select PFC neural ensembles and their connections to limbic and amygdalar regions play a large role in the regulation of social behaviors.

Given our previous sex-specific findings of AIE-associated alteration of the OXT and AVP systems, it is also possible that regions with known sex differences in these neuropeptide systems are differentially involved in modulation of social interaction in AIE-exposed males relative to controls as well as differences between control conditions in each sex. For example, the LS receives greater AVP innervation from the bed nucleus of the stria terminalis (BNST) (De Vries et al., 1981; De Vries and Buijs, 1983) and HYPO (Qiao et al., 2014) in male than in female rodents. Sex differences in the expression of OXT and AVP receptors are also evident across multiple brain regions, including the PFC, subregions of the BNST, NAc, medial amygdala (MeA), and ventromedial HYPO (Dumais et al., 2013; Dumais and Veenema, 2016). In addition, AVP synthesis in regions such as the BNST has been shown to be different between sexes (Van Leeuwen, Caffe, and De Vries, 1985). However, it is still unknown whether neuronal activation within regions that show sex differences in the OXT and AVP systems in response to a social stimulus differs as a function of adolescent ethanol exposure and sex, and whether these brain regions contribute to AIE-induced social alterations in males.

We hypothesized that ethanol exposure during adolescence may affect the brain in a sex-specific manner, with brain regions implicated in social behavior regulation, especially those that show sex differences, being sensitive to AIE in males. The current study was designed to assess neuronal activation induced by a social stimulus in “social” brain regions of adult male and female cFos-LacZ rats exposed to AIE. In these transgenic rats, expression of β-galactosidase (β-gal) is under the control of a cfos promoter (Figure 1B), and cFos and β-gal proteins are co-expressed in neurons strongly activated by a certain stimulus (Cruz et al., 2015). Therefore, β-gal expression can be used as a proxy for cFos. While cFos expression requires immuno-histochemical techniques, β-gal is detected by X-gal staining–a rapid and highly reproducible histochemical assay. In this study, β-gal expression was assessed in the following regions of interest (ROIs): prelimbic cortex (PrL), infralimbic cortex (IL), ventral and lateral orbitofrontal cortex (vOFC, lOFC), nucleus accumbens core (NAcC), nucleus accumbens shell (NAcSh), lateral septum (LS), dorsal bed nucleus of the stria terminalis (dBNST), anterior hypothalamus (AH), central amygdala (CeA), and basolateral amygdala (BLA).

In this study, we aimed to determine brain regions involved in modulation of social behavior that might contribute to AIE-induced social alterations. Similar to our previous findings in non-transgenic Sprague Dawley rats (Varlinskaya et al., 2014; Varlinskaya et al., 2017; Dannenhoffer et al., 2018; Varlinskaya et al., 2020), AIE led to a sex-specific social impairment in transgenic cFos-LacZ animals, with only males displaying a decreased social preference following adolescent ethanol exposure.

The analyses of relationships between social behaviors and neuronal activation of the ROIs revealed that these relationships are altered by AIE, with these alterations evident in both males and females. Specifically, in water-exposed control males, frequency of social investigation was found to be positively correlated with the number of β-gal + neurons in the PrL, IL, and NAcC, suggesting an involvement of these brain regions in regulation of male social investigation. In contrast, no relationship between activation of these regions and social investigation was evident in AIE-exposed males, suggesting AIE-induced alterations of the neural circuitry involved in modulation of social investigation. Alterations of the mPFC and NAc induced by AIE exposure may limit the ability of these regions to contribute to social investigation. Identified AIE-induced changes in mPFC inhibitory signaling include reduced interneuron intrinsic excitability (Trantham-Davidson et al., 2017), increased inhibitory signaling onto layer V pyramidal neurons (Centanni et al., 2017), and decreases in the number of parvalbumin expressing interneurons (Rice et al., 2019), with these changes to inhibitory signaling in this region potentially contributing to a dysregulated activation associated with social investigation. Additionally, AIE has been found to blunt stimulus evoked dopamine signaling in the NAc (Shnitko et al., 2016) and potentiate kappa opioid receptor function in mediating dopamine release (Spodnick et al., 2020). All these alterations might also contribute to male-specific AIE-associated changes observed in the present study.

Activation of mPFC and NAcC of females did not contribute to social investigation in both adolescent exposure conditions, suggesting sex-specific involvement of these regions in social investigation and AIE-induced deficits. In contrast to males, social investigation in water-exposed females was negatively correlated with vOFC neuronal activation, with no such relationship evident in AIE females, suggesting AIE-associated neuronal alterations in modulation of social behavior in females. As mentioned above, the OFC modulates various social behaviors in males, however the involvement of this region in female social investigation has not been thoroughly studied. It was surprising to find a negative relationship between OFC activity and female social investigation in the current study, given that OFC lesions reduced social play (Van Kerkhof et al., 2013) and increased aggression (Rudebeck et al., 2007). However, the involvement of the OFC in social behaviors has not been evaluated in females, therefore it is likely that regulation of social investigation by the OFC differs between females and males. Sex differences in the OFC are also evident, with females having more myelin basic protein (Bayless and Daniel, 2015; Darling and Daniel, 2019) and greater OFC volume than males (Goldstein et al., 2001; Gur et al., 2002), although the contribution of these sex-differences to social behavior remain unknown.

The relationship between social preference and neuronal activation differed from that of social investigation. In water-exposed control males, β-gal expression in the dBNST was positively correlated with social preference, while AIE reversed this relationship, suggesting AIE-associated alterations in the role of the dBNST in social preference. Indeed, the BNST has been shown to assess social contexts in order to select an appropriate behavioral response, while alterations of socially relevant BNST systems by different insults result in social dysfunctions (Flanigan and Kash, 2020). In addition, social preference was negatively correlated with activation of the vOFC and LS in AIE-exposed males, whereas it was negatively correlated with activation of the NAcC and NAcSh in AIE-exposed females. These findings suggest that AIE produces sex- and region-specific dysfunctions in brain regions implicated in social preference/avoidance. In other words, activation of brain regions that are normally implicated in social behavior (Vanderschuren et al., 2016; Menon et al., 2021) is associated with lower social preference and/or social avoidance following AIE, potentially due to AIE-induced changes in either expression of certain cell types expressing β-gal or alterations to connectivity, and/or neurotransmitter release within these regions compared to control animals. Clearly, more work is needed to understand neural mechanisms of dysfunctions in brain regions implicated in social preference/avoidance sex-specifically affected by AIE. Furthermore, the results of the present study suggest that social investigation and social preference of males have different underlying circuitry, with IL, PrL, and NAcC implicated in regulation of social investigation and BNST contributing to social preference. This finding was not surprising, given that social investigation is a form of exploratory behavior that provides needed information about a novel conspecific, whereas the coefficient of social preference/avoidance reflects approach/avoidance behavior under social circumstances.

When assessing neuronal activation following an interaction with a social partner and under home cage condition in cFos-LacZ rats using β-gal as the proxy for cFos, we found that AIE and water-exposed males had differential activation of the CeA regardless of testing condition, with AIE-exposed males having less β-gal + cells in the CeA than water-exposed controls. It is likely that in males, AIE produces a long-lasting reduction in CeA activity indexed through β-gal expression. This finding is in accordance with others that have shown AIE-associated alterations in the CeA, including decreases expression the activity-regulated cytoskeletal (Arc) gene (Pandey et al., 2015; Kyzar et al., 2019), another early gene marker used to assess neuronal activation (Korb and Finkbeiner, 2011). We cannot preclude, however, that the attenuated activation state of the CeA in AIE males plays a role in AIE-induced social impairments by predisposing these males to avoidance behavior. Future studies are needed to further assess the role of the CeA in social deficits associated with AIE. Although no difference in CeA neuronal activation between exposure conditions was evident in socially tested females, we found that β-gal expression in the CeA under home cage conditions was also affected by AIE. However, in contrast to males in this study and others (Pandey et al., 2015; Kyzar et al., 2019), AIE-exposed females had more β-gal + cells in the CeA than their water-exposed counterparts. Interestingly, social behavior led to a recruitment of neurons in the CeA of control females, however this effect was absent in AIE-exposed females. Although no behavioral difference was observed between female exposure conditions, it is possible that social behavior in each of these exposure groups relies on differential activation of brain regions such as the CeA. These findings add to the growing literature suggesting that AIE leads to sex-specific alterations (Robinson et al., 2021), including opposite effects of AIE on CeA in males and females.

In females, social investigation and social preference did not differ as a function of adolescent exposure, however, the OFC was differentially activated in water- and AIE-exposed females. In, ventral and lateral subregions of the OFC, AIE-exposed females had lower β-gal expression than their water-exposed counterparts, with these differences evident in both home cage controls and socially tested subject, indicating AIE-associated alterations of the OFC in females. Alterations of the OFC following AIE have been recently identified. For example, AIE leads to reduced cholinergic markers and lower acetylcholine release in the OFC during a spontaneous alternation task, although these results were similar between sexes (Kipp et al., 2021). AIE also results in increased neuroimmune signaling in the OFC of males (Vetreno et al., 2013), however this was not assessed in females. Furthermore, AIE has been shown to increase perineuronal net (PNNs) expression in the OFC (Coleman et al., 2014) and given that these nets mediate inhibitory signaling (Bosiacki et al., 2019; Carceller et al., 2020), it is possible that PNN upregulation contributes to the increased neuronal activity among AIE-exposed females evident in the current study. The increases in neuroinflammation and perineuronal nets following AIE have only been studied in males, and more work is needed to determine whether these changes are evident in females and whether they contribute to differences in basal activity of the OFC. In both sexes, however, exposure to a social stimulus enhanced neuronal activation of the vOFC and lOFC, with greater number of β-gal-containing neurons evident in socially tested animals than home cage controls regardless of adolescent exposure condition. This finding suggest that the OFC contributes to social responding of adult males and females, and that the OFC is still responsive to social stimuli after AIE. Evidence supports a role of the OFC in modulating different social behaviors, including social play (Van Kerkhof et al., 2013), aggression (Rudebeck et al., 2007), and dominance (Kolb et al., 2004), whereas no information is available regarding possible involvement of the OFC in regulation of social investigation and/or social preference in males and females.

The present study has certain limitations. Our initial analyses using independent samples t-tests revealed significant differences between socially tested adolescent exposure conditions in the vOFC and lOFC however when using two-way ANOVAs these comparisons failed to reach significance. It is likely that our effect was small in these analyses and our power was too low for identification of these differences with additional analyses. Future work will determine whether these differences in OFC activation between socially tested females can be replicated. Moreover, a lack of differences in activation of a certain brain region does not imply that the same types of neurons were activated by a social stimulus in water-exposed controls and AIE animals. Therefore, our future studies will focus on identification of subtypes of neurons activated by exposure to a social partner in ROIs activation of which was correlated either with social preference or social investigation. In addition, the use of a single immediate early gene limits our understanding of the timeframe in which a neuron was active. The use of either different immediate early genes or different timepoints may be appropriate to further assess the activation of the ROIs in relation to social behavior. Furthermore, we cannot rule out that changes associated with social interaction relative to home cage levels of β-gal expression can be detected in other ROIs examined, since baseline neuronal activation was only assessed in regions with significant differences in activation between socially tested water- and AIE-exposed males and females. Future studies will determine whether AIE induces alterations in basal β-gal expression and whether exposure to a social stimulus produces changes relative to baseline condition differentially in AIE animals and water-exposed controls.

Overall, in males, AIE reduced social preference and affected the relationship between social behaviors’ and neuronal activation of IL, PrL, vOFC, NAcC, BNST, and LS. Importantly, although no behavioral alterations were observed in AIE-exposed females, our findings suggest that the relationship between social behaviors and neuronal activation of vOFC, nAcC, and NAcSh were affected by AIE. Further studies need to be conducted to determine neuronal phenotypes are activated in response to a social stimulus and to test whether these phenotypes differ as a result of AIE exposure in males and females.