A female's behavioral and physiological response to her environment can have long-lasting effects on the phenotype of her developing offspring, potentially providing a way to prepare them for the current environment (Mousseau, 1998; Uller, 2008). For example, hormone-mediated maternal effects, in which maternal hormones prenatally influence offspring phenotype, are seen across taxa, including fish (McCormick, 1999), reptiles (Uller et al., 2007), mammals (Dantzer et al., 2011; Dloniak et al., 2006), and insects (Crocker and Hunter, 2018). However, birds tend to be the best studied due to the relative ease of manipulating and measuring hormones in their externally developing embryos (Groothuis et al., 2005). One well-studied hormone-mediated maternal effect occurs in response to the social environment. In many species, females breeding in more competitive environments lay eggs with higher concentrations of yolk testosterone (T) (Bentz et al., 2013; Bentz et al., 2016a; Hargitai et al., 2009; Mazuc et al., 2003; Navara et al., 2006a; Pilz and Smith, 2004; Schwabl, 1997; Whittingham and Schwabl, 2002). Offspring that are exposed to elevated yolk T experience numerous phenotypic changes, including faster juvenile growth (Bentz et al., 2013; Eising et al., 2001; Navara et al., 2006b; Pilz et al., 2004; Schwabl, 1996), modified immune function (Navara et al., 2005; Navara et al., 2006b), and increased aggressive behaviors that persist into adulthood (Bentz et al., 2021a; Eising et al., 2006; Partecke and Schwabl, 2008; Strasser and Schwabl, 2004). Despite the wide-ranging and potentially adaptive effects of maternally derived steroids on offspring, the mechanisms causing these phenotypic changes remain unclear (Groothuis and Schwabl, 2008).

Few studies have examined the molecular mechanisms that underlie T-mediated maternal effects (Groothuis et al., 2019). Past work has primarily focused on neural sex steroid receptor genes, like estrogen (ER; Bentz et al., 2016b) and androgen receptors (AR; Pfannkuche et al., 2011), as candidates that could be sensitive to T and mediate the pleiotropic effects observed in offspring, including aggressive behaviors (Nelson and Trainor, 2007) and growth (Chang et al., 1995; Nilsson and Gustafsson, 2002). Yet, evidence from early embryonic stages suggests that sex steroid receptor expression in the head is not responsive to experimentally elevated yolk T (Kumar et al., 2019a). Steroid receptors likely play a more nuanced role, e.g., through receptor-mediated effects on downstream genes (Brinkmann et al., 1999) or non-genomic actions (Foradori et al., 2008). However, we need more information on how these processes interact with maternal hormones during development, because it is increasingly evident that maternally derived T acts through more diverse mechanisms than previously explored. For example, one of the best-studied mechanisms linking early life exposure to sex steroids and later adult traits involves immune signaling. Sexual differentiation of the brain causes lasting behavioral changes through interactions of sex steroids, inflammatory signals, and the resident immune cells of the brain (microglia; Arambula and McCarthy, 2020; Delage and Cornil, 2020; Nelson and Lenz, 2017). Microglia can shape neural development and synaptic connectivity, leading to lasting changes in neural function (VanRyzin et al., 2020), and are associated with other early life experiences that impact later-life cognition and behavior (e.g., maternal infection, Bilbo and Schwarz, 2009; prenatal stress, Gómez-González and Escobar, 2010). Additionally, T metabolites may also play a role. Maternally derived T is rapidly metabolized early in incubation to etiocholanolone (ETIO; Campbell et al., 2020; Kumar et al., 2019b), which does not bind to AR (Fang et al., 2003) but can still affect embryo development (Irving et al., 1976; Levere et al., 1967; Wang et al., 2023a; Wang et al., 2023b). Thus, narrowly focusing on transcriptional changes in classical nuclear receptors leaves uncertainty about the diverse processes potentially underlying hormone-mediated maternal effects. A transcriptome-wide approach during the critical window of exposure would help to clarify these mechanisms.

Here, we explore the effects of the maternal social environment on yolk T concentrations and genome-wide patterns of neural gene expression in tree swallow (Tachycineta bicolor) embryos. We predicted that if the maternal environment elevates yolk T, then genes involved in steroid synthesis, signaling, or metabolism, as well as neuroimmune processes, are likely candidates to be affected. Tree swallows are a good model for social maternal effects as females are highly territorial (Rosvall, 2008) and show transcriptomic (Bentz et al., 2021b; Bentz et al., 2022) and hormonal changes (George et al., 2022) in response to competition. Furthermore, elevated yolk T is found in response to competition, including elevated breeding density (Bentz et al., 2013) and natural territorial intrusions (Whittingham and Schwabl, 2002), and offspring exposed to elevated yolk T show faster growth and enhanced competitive ability as juveniles (Bentz et al., 2013). In this study, we observed a population of free-living tree swallows breeding in nest boxes at sites with variable breeding densities. We collected eggs at two timepoints, including the day laid to measure yolk T concentrations and eggs incubated to embryonic day (ED) 11 to measure gene expression in whole brains with RNA-seq. We assessed patterns of yolk T allocation across breeding density and determined the degree to which the maternal social environment relates to patterns of gene expression in the embryonic brain. Embryos included both males and females because yolk T can have sex-specific effects on growth, begging rate, and survival (Ruuskanen and Laaksonen, 2010; Sockman et al., 2008). Our data highlight several key genes associated with variation in breeding density and yolk T that point to potential mechanisms by which the maternal social environment could impact phenotypic plasticity in offspring.