Mammalian species show a wide diversity of social behaviors. This natural variation can be leveraged to identify the unique circuitry that imparts a behavior of interest. Several vole species are established model organisms for the study of sociality (Carter et al., 1995; He et al., 2019; Insel and Young, 2001; Lee and Beery, 2019; Sadino and Donaldson, 2018). There are stark behavioral differences between monogamous and promiscuous vole species: prairie voles display social monogamy, biparental care, and form long-lasting pair bonds, which rarely occur in mammals, while promiscuous meadow voles are socially non-monogamous, uniparental, and seasonally social in same-sex groups (Anacker et al., 2016; Beery, 2019). Notably, social monogamy is accompanied by decreased behavioral sex differences in prairie voles compared to mice and rats. Male prairie voles care for offspring and juveniles and show high levels of spontaneous alloparenting (Bales et al., 2006; Carter and Getz, 1993; Getz et al., 1981; Kramer et al., 2009; Lonstein and De Vries, 2000). Prairie voles also display similar levels of aggression in females and males. Males attack novel conspecifics at low levels, but upon pair-bond formation, both sexes show selective aggression towards strangers (Getz et al., 1981; Lee and Beery, 2022; Tickerhoof et al., 2020; Wang et al., 1997; Young et al., 2011). Thus, a pair-bond encompasses both prosocial (towards the partner) and antisocial (towards novel conspecifics) behaviors.

Social behaviors, such as aggression, mounting, lordosis, and parenting behaviors, are mediated by the social behavior network (SBN), which processes pheromonal cues to determine social context and select the appropriate response (Chen and Hong, 2018; Goodson, 2005; Newman, 1999; Wallace et al., 2023). In many species, including humans, several regions of the SBN show sexual dimorphism in cell density, cell number, volume, projection pattern, or gene expression, particularly the sexually dimorphic nucleus of the preoptic area (SDN-POA), the principal nucleus of the bed nucleus of the stria terminalis (BNSTpr), and anteroventral periventricular nucleus (AVPV) (Allen and Gorski, 1990; Gorski et al., 1978; Kelly et al., 2013; Simerly et al., 1985; Tsukahara and Morishita, 2020). In mice and rats, these dimorphisms are organized by perinatal testosterone signaling in males. Circulating testosterone is locally converted to 17β-estradiol in select neuronal populations that express aromatase and this neural estradiol drives brain sexual differentiation (Balthazart and Ball, 1998; Juntti et al., 2010; Lephart, 1996; MacLusky and Naftolin, 1981; Naftolin and Ryan, 1975; Wu et al., 2009). Intriguingly, monogamous species such as prairie voles also show decreased sexual dimorphism in the brain, as well as in anatomical measures such as body weight and anogenital distance (Campi et al., 2013; Dewsbury et al., 1980; Heske and Ostfeld, 1990; Shapiro et al., 1991). In addition, prairie vole social behaviors are resistant to early life testosterone manipulations, as if the classic organization and activation model is operating under different constraints in this species. Like other rodents, prairie vole males undergo a perinatal testosterone surge (Lansing et al., 2013), but in contrast to rats, neonatal orchiectomy does not abolish adult male sexual behavior. Instead, postnatal testosterone reduces androgen induced mounting behavior in adult males (Roberts et al., 1997). Perinatal testosterone also does not masculinize the expression of arginine vasopressin (Avp) in the BNST and medial amygdala (MeA) (Lonstein et al., 2002). These findings suggest that prairie vole brain sexual differentiation follows a distinct trajectory compared to those of mice and rats (Bonthuis et al., 2010).

We reasoned that behaviors associated with social monogamy may emerge in part from unique expression patterns of gonadal steroid hormone receptors in prairie voles. Previous studies have investigated the distribution of ERα immunoreactivity across group-housed vole species in adulthood, demonstrating differences in overall ERα levels within the SBN, as well as decreased expression in males of socially-monogamous species, particularly in the MeA (Cushing et al., 2004; Cushing and Wynne-Edwards, 2006; Hnatczuk et al., 1994). However, male mice and rats also show lower expression of ERα/Esr1 in males compared to females (Cao and Patisaul, 2013; Gegenhuber et al., 2022; Kanaya et al., 2018; Kelly et al., 2013; Xu et al., 2012; Yokosuka et al., 1997), and no studies have directly compared hormone receptor expression between voles and mice. In addition, although immunoreactivity to ERβ has been reported to be less sexually dimorphic than ERα in prairie voles (Ploskonka et al., 2016), AR expression in this species has only previously been described in males (Cushing et al., 2004).

To obtain insight into species differences in brain sexual differentiation, we carried out a systematic comparison of Esr1, Esr2, and Ar, using in situ hybridization (ISH) at postnatal day 14 (P14) in mice, prairie, and meadow voles. The inclusion of meadow voles permits comparisons within two Microtus species and provides a behavioral intermediate between polygamous mice and monogamous prairie voles: meadow voles show a partner preference for same-sex peers and engage in seasonal social group living (Anacker et al., 2016; Beery, 2019; Beery et al., 2014; Lee and Beery, 2022). We selected ISH to allow a consistent methodology across all three genes. Immunostaining for ERβ has been historically complicated by antibody variability; a previous comprehensive characterization of ERβ expression in mice utilized a transgenic Esr2-GFP line (Andersson et al., 2017; Nelson et al., 2017; Zuloaga et al., 2014), while experiments in prairie voles utilized an antibody raised against a prairie vole epitope (Ploskonka et al., 2016). We chose P14 as this time point is beyond the closure of the postnatal sensitive period for brain sexual differentiation in mice and rats (MacLusky and Naftolin, 1981; McCarthy, 2008; Simerly, 2002), precedes pubertal gonadal hormone secretion (Brock et al., 2011; Piekarski et al., 2017). P14 also coincides with a dynamic increase in the expression of the oxytocin receptor (Oxtr), a gene that underlies species differences in social behavior (Hammock and Levitt, 2013; Hammock, 2014; Insel et al., 1993; Johnson and Young, 2015; Newmaster et al., 2020). Since prairie voles in particular are more precocial in their early postnatal development than promiscuous species (Shapiro and Insel, 1990) in terms of tooth eruption, fur growth, and eye opening, we anticipated that the closure of the sensitive period in voles would be at an earlier postnatal age as in mice (or at the same time, but not a later age). We additionally investigated sex differences in hormone receptor expression in the postnatal BNSTp.