Elsevier

Hormones and Behavior

Volume 106, November 2018, Pages 178-188
Hormones and Behavior

Neural activity in the social decision-making network of the brown anole during reproductive and agonistic encounters

https://doi.org/10.1016/j.yhbeh.2018.06.013Get rights and content

Highlights

  • A Social Decision-Making Network (SDMN) is present in reptiles.

  • SDMN activity varies with social context, but generally not with specific behavioral output.

  • Specific SDMN nodes do correlate with behavioral intensity and frequency.

  • Functional connectivity of the SDMN is present during social encounters and greatest in a reproductive context.

  • SDMN activity correlates with oxytocin activity, and inversely with vasopressin, catecholamine, and serotonin activities.

Abstract

Animals have evolved flexible strategies that allow them to evaluate and respond to their social environment by integrating the salience of external stimuli with internal physiological cues into adaptive behavioral responses. A highly conserved social decision-making network (SDMN), consisting of interconnected social behavior and mesolimbic reward networks, has been proposed to underlie such adaptive behaviors across all vertebrates, although our understanding of this system in reptiles is very limited. Here we measure neural activation across the SDMN and associated regions in the male brown anole (Anolis sagrei), within both reproductive and agonistic contexts, by quantifying the expression density of the immediate early gene product Fos. We then relate this neural activity measure to social context, behavioral expression, and activation (as measured by colocalization with Fos) of different phenotypes of ‘source’ node neurons that produce neurotransmitters and neuropeptides known to modulate SDMN ‘target’ node activity. Our results demonstrate that measures of neural activation across the SDMN network are generally independent of specific behavioral output, although Fos induction in a few select nodes of the social behavior network component of the SDMN does vary with social environment and behavioral output. Under control conditions, the mesolimbic reward nodes of the SDMN actually correlate little with the social behavior nodes, but the interconnectivity of these SDMN components increases dramatically within a reproductive context. When relating behavioral output to specific source node activation profiles, we found that catecholaminergic activation is associated with the frequency and intensity of reproductive behavior output, as well as with aggression intensity. Finally, in terms of the effects of source node activation on SDMN activity, we found that Ile8-oxytocin (mesotocin) populations correlate positively, while Ile3-vasopressin (vasotocin), catecholamine, and serotonin populations correlate negatively with SDMN activity. Taken together, our findings present evidence for a highly dynamic SDMN in reptiles that is responsive to salient cues in a social context-dependent manner.

Introduction

All animals encounter challenges (e.g., territorial intrusions; competition for mates) as well as opportunities (e.g., finding a mate; ascending to social dominance) throughout their lives. Their behavioral decisions in these situations fundamentally affect their chances of survival and reproduction (O'Connell and Hofmann, 2011a; Rittschof and Robinson, 2016). O'Connell and Hofmann (2011b) proposed that a relatively conserved social decision-making network (SDMN; see Table 1 for list of abbreviations), consisting of overlapping social behavior and mesolimbic reward system neural networks, underlies these decision-making processes, allowing vertebrates to evaluate the salience and valence of stimuli in relation to their life history. Changes in neural activity at nodes within this network are associated with corresponding changes in behavioral expression. This nodal activity is modulated by a number of signaling molecules (e.g., nonapeptides, catecholamines, indolamines, steroid hormones) that have been shown to regulate social behaviors on an individual basis, and differently within different social contexts (Crews, 2003; Goodson, 2005; Goodson and Kabelik, 2009; Newman, 1999; O'Connell and Hofmann, 2012; Yang and Wilczynski, 2002). Furthermore, the functional connectivity of nodes within the network can be influenced by social environment, at least partly via signaling molecules – e.g., via the nonapeptide oxytocin during reproductive opportunities (Johnson et al., 2016). However, much remains unknown about how various other social environments, and different signaling molecules originating from and targeting different brain regions, affect neural activity across the multitude of nodes of the SDMN. Even whether a functional SDMN is present across all vertebrate groups has been questioned (Goodson and Kingsbury, 2013). We here present evidence for a functional SDMN in reptiles, and examine neural activity differences, as well as differences in functional connectivity to and within this network, across and within Reproductive (including courtship and intromission), Agonistic (relating to conflict), and Control (solitary, non-social) environments in the male brown anole lizard (Anolis sagrei).

While birds and mammals are well-researched in the field of behavioral neuroscience, reptiles remain relatively overlooked; in fact the study of behavioral neuroscience in reptiles also lags that of amphibians, fish, and invertebrates (Kabelik and Hofmann, this issue; Taborsky et al., 2015). This lack of research on reptiles is a significant problem for understanding the evolution and function of the vertebrate SDMN, given reptiles' important position in vertebrate phylogeny. Examination of lizards is important for evolutionary comparisons among amniotic vertebrates (mammals, birds, and reptiles). Mammals evolved from reptile-like synapsids, while all other extant amniotes evolved from diapsids, with lizards and other squamates constituting the majority of the Lepidosauria, a separate group from the Archelosauria (Archosauromorpha) which comprise birds, crocodiles, and likely also turtles (Crawford et al., 2015).

By examining a small vertebrate brain (that of the brown anole lizard), and implementing immunofluorescent detection of chemoarchitecture and a marker of neural activity (the production of the immediate early gene product Fos), we here present an analysis of neural activity at various source signaling molecule cell clusters (‘source’ nodes), as well as at ‘target’ SDMN nodes that are regulated by the neurotransmitters and neuropeptides originating from the source nodes.

The target nodes examined in the present study primarily include brain nuclei that have been specified as part of the SDMN (O'Connell and Hofmann, 2011b) (Fig. 1). The nodes include components of the social behavior network such as the preoptic area (POA), anterior hypothalamus (AH), ventromedial nucleus of the hypothalamus (VMH) and periaqueductal gray (PAG; a.k.a., central gray), as well parts of the mesolimbic reward system such as the nucleus accumbens (NAc), the basolateral amygdala (AMYbl, often described in reptiles as the posterior dorsal ventricular ridge but a presumed homologue of the mammalian AMYbl, (Lanuza et al., 1998; Martínez-García et al., 2002)), and the ventral tegmental area (VTA). Examined brain regions that are part of both networks include the lateral septum (LS) and the bed nucleus of the stria terminalis (BNST)/medial amygdala (AMYm) continuum. Additional regions examined within this study include the ventromedial septum (VMS), which is a distinct septal region in lizard brains that possesses strong connectivity with the AH as well as with the BNST and parts of the amygdala (Font et al., 1997, Font et al., 1998). We also examine the cortical amygdala (AMYc, a.k.a., ventral anterior amygdala, which is thought to be homologous to the mammalian anterior and posterolateral cortical amygdala, Martínez-García et al., 2002), and the paraventricular nucleus of the hypothalamus (PVN; implicated in stress responsiveness, Goodson and Kabelik, 2009; Goodson and Kingsbury, 2013). Furthermore, we examined the substantia nigra (SN) and the location of the A8 catecholamine population, both additional dopaminergic midbrain sites that are part of the mesolimbic reward system along with the VTA. We also examined activity in the torus semicircularis (TS), as it is continuous with the PAG and possesses some behavioral functions. Although parts of the TS are at least partly homologous to the inferior colliculus (Butler and Hodos, 2005), and thus likely possesses some auditory function, the TS also possesses other roles that make it important for inclusion in the SDMN. For instance, the TS is an intermediary relay area between the LS and the PAG (Font et al., 1998). Sex steroid hormone receptors are also present in the TS of anoles (Morrell et al., 1979), and electrical stimulation of the TS in the green iguana (Iguana iguana) has been shown to elicit dewlap displays (Distel, 1978). The TS regions sampled here have also been shown to express substance P, as does the lateral PAG of mammals (Goodson and Kingsbury, 2013), raising the possibility that these regions are also partly homologous.

In the present study, the signaling molecules that were visualized directly include the lizard vasopressin (VP; Ile3-vasopressin, a.k.a. Arg8-vasotocin) and the lizard oxytocin (OT; Ile8-oxytocin, a.k.a. mesotocin). We use the terms vasopressin and oxytocin here, rather than vasotocin and mesotocin, according to the example of Kelly and Goodson (2014), to denote homology with the mammalian forms of the nonapeptides from which they vary solely by one amino acid substitution; in mammals, equivalent differences do not constitute deviations from this nomenclature (e.g., Arg8-vasopressin and Lys8-vasopressin are both referred to as vasopressin, while Leu8-oxytocin and Pro8-oxytocin are both referred to as oxytocin) (Albers, 2015; Lee et al., 2011). We also directly examined neurons producing the indolamine serotonin (5-hydroxytryptamine, 5-HT). Furthermore, we examined neurons producing the catecholamines dopamine (DA), norepinephrine (NE), and epinephrine (Epi) by means of visualizing the rate-limiting enzyme in catecholamine synthesis, tyrosine hydroxylase (TH). Neural activity within these signaling molecule-producing neurons, in relation to social behavior context and behavioral expression in brown anoles, was previously examined separately by means of immunofluorescent colocalization with Fos (Hartline et al., 2017; Kabelik et al., 2013, Kabelik et al., 2014; Kabelik and Magruder, 2014). Fos is a robust marker of neural activity, and its upregulation is induced both by large-scale neural depolarization and by non-action potential-coupled pathways such as calcium influx due to receptor activation (Sabatier et al., 2003). In the present study, we examine brain sections from these same male brown anoles, but now focus primarily on target SDMN node activation, correlations among the nodes, and how nodal activity and coordination is influenced by source node activity and social context. Specifically, we expose A. sagrei males to either a conspecific female (Reproductive opportunity) or a conspecific male (Agonistic challenge) and contrast these treatments to a solitary Control context. Importantly, the Control context still exposes focal males to a novel environment, as in the experimental conditions, but there is no social stimulus present in this novel environment. We predicted that target node activity within the SDMN will generally increase in social versus asocial contexts, but the opposite was found. We also expected and did find correlated activation of select SDMN nodes with increased social behavior expression. Furthermore, we predicted increased functional connectivity (correlated activity) among SDMN nodes in social behavior contexts relative to the Control condition, and found this true for the Reproductive context. Finally, we expected source node activity to correlate in both positive and negative directions with target node activity (depending on the specific brain nucleus and signaling molecule examined), and we did find such results, with oxytocin connections to the SDMN showing a positive correlation while other regulatory molecules correlated negatively to general SDMN activity.

Section snippets

Subjects and treatment groups

Fifty-seven adult male brown anoles (Anolis sagrei) served as experimental subjects and were divided into treatment groups of Control (no conspecific present, N = 12), Reproductive opportunity (female conspecific present, N = 22), and Agonistic challenge (male conspecific present, N = 23). Data obtained from these same animals regarding Fos colocalization with various signaling molecules have previously been reported (Hartline et al., 2017; Kabelik et al., 2013, Kabelik et al., 2014; Kabelik

Analysis of aggression and courtship behavior within treatments

We first examined the latency, frequency, and intensity of aggressive (Agonistic challenge) and courtship (Reproductive opportunity) displays (Supplementary Fig. 2). Because all three behavioral measures were strongly correlated (see Spearman correlation coefficients and p-values in Supplementary Table 1), we used behavioral frequencies as the sole measure in analyses correlating behaviors to Fos-ir density measurements within various brain nuclei.

Between-treatment and within-treatment analyses of SDMN activity (Fos-ir density)

Next, we examined Fos-ir densities in SDMN

Discussion

In the present study, we examined neural activation patterns across the SDMN and associated regions in the male brown anole, Anolis sagrei, within both Reproductive and Agonistic contexts. We then related this neural activity measure to the activity (as determined by colocalization with Fos) of different ‘source’ nodes that release neurotransmitters and neuropeptides that can modulate SDMN ‘target’ node activity. Overall, our results show that: (1) neural activity of a core network of SDMN

Acknowledgements

We would like to thank Rhodes College, as well as Dr. Charles and Mrs. Patricia Robertson, for their generous support of DK, SCC, JTH, and ANS. This work was supported by an NSF Graduate Research Fellowship to CAW and NSF grants IOS-1501704 to HAH and IOS-1601734 to CAW and HAH.

References (39)

Cited by (24)

  • A consideration of brain networks modulating social behavior

    2022, Hormones and Behavior
    Citation Excerpt :

    Lastly, I would argue that there is a lack of a control for testing that the SBN is indeed “the core” of the social brain as suggested by Goodson (2005). Several studies examine how neural responses vary across the SBN between species or treatment groups (Goodson et al., 2005; Maney et al., 2008; Kabelik et al., 2018; Petersen et al., 2021). Yet, one could pick any number of brain regions at random and expose animals to different contexts and see different “patterns” of activity across the brain regions simply because, individually, the distinct brain regions may differentially respond to varying contexts.

  • Aggressive but not reproductive boldness in male green anole lizards correlates with baseline vasopressin activity

    2022, Hormones and Behavior
    Citation Excerpt :

    Vasopressin is a neuromodulator that has previously been shown to have causal effects on aggression (Kelly and Goodson, 2014b; Terranova et al., 2017), and our study supports the notion that basal VP activity helps to determine individual differences in aggressive behavior output. Our results are in line with previous findings showing a negative correlation between neuronal VP activity and activation of social decision-making network nodes in the closely related brown anole, especially within agonistic contexts (Kabelik et al., 2018). Interestingly, although boldness measures are correlated between agonistic and reproductive contexts, VP activity does not correlate well with reproductive boldness suggesting that other mechanisms may underlie the cross-context correlation.

  • Social boldness correlates with brain gene expression in male green anoles

    2021, Hormones and Behavior
    Citation Excerpt :

    Likewise, these groups did not differ in testes mass, or in circulating testosterone, estradiol, progesterone, or glucocorticoid levels. We measured baseline gene expression in bold and shy individuals across six brain regions that contribute to social decision-making and are functionally conserved across vertebrates (Kabelik et al., 2018; Newman, 1999; O'Connell and Hofmann, 2011, 2012; Thompson et al., 2008; Walton et al., 2010) (Fig. 2A; Supplemental Excel File). We used a principal component analysis to visualize overall gene expression differences across brain regions and groups.

  • Differences in cautiousness between mainland and island Podarcis siculus populations are paralleled by differences in brain noradrenaline/adrenaline concentrations

    2020, Physiology and Behavior
    Citation Excerpt :

    As behavior is primarily regulated through an interaction between hormones and neural circuits, central monoaminergic systems are expected to play an important role in mediating the response to a stressor. Changes in brain catecholamine and 5-HT concentrations have been reported following stressful events such as aggressive interactions, social challenge, and forced physical activity or restraint [26,38,54,57], and catecholaminergic activation was shown to be associated with the frequency and intensity of reproductive behavior and aggression [24]. We previously compared monoaminergic profiles between the two Podarcis species, and found significantly higher dopamine levels in P. siculus, which can outcompete P. melisellensis when they are found in syntopy [34].

View all citing articles on Scopus
View full text