Sexual coordination in a whole-brain map of prairie-vole pair-bonding

Reviewer #1 (Public Review):

This is an intriguing and creative paper that examines whole brain cfos induction, a measure of brain activity, during mating and the formation of pair bonding. This contrasts with the classical reductionistic approach of focusing on a few individual brain regions in the monogamous and pair-bonding prairie vole. By taking this whole-brain approach following mating and then bond formation, several findings are revealed. (1) Using hierarchical cluster analysis some clusters were consistent with previously well-identified brain regions/circuits involved in bonding. The bed nucleus of the stria terminalis was identified as an important hub for bonding behavior but, importantly, the study also identified newer brain circuits likely involved in pair bond maintenance. (2) Rates of ejaculation best predicted the consistency of cfos activation that characterized a pair. (3) once the pair bond has stabilized, an amygdala cluster emerged potentially representing the coordination of a new cluster of brain regions that allow for pair bond maintenance. (4) There was a surprising lack of sexual dimorphism in active brain clusters identified during mating and pair bonding, but perhaps characteristic of a monogamous species.

While the approach used in this study cannot identify cause and effect, the whole brain approach identified clusters representing circuits of potential importance and a series of new hypotheses to explore.

The importance of the role of sexual behavior, specifically ejaculation rates, is worth emphasizing for the formation of pair bonds in prairie voles. It suggests that the role of sexual behavior in contributing to the strength of pair bonds should be explored more. It is also important to add that males and females in the study were screened for sexual receptivity. It would therefore be important to identify characteristics of animals that did not mate under the laboratory conditions used that may add depth and complexity to what was identified in the current study. The identification of brain regions for pair bond maintenance centered around the amygdala was also intriguing.

The issue of the lack of a strong presence of the reward circuitry (nucleus accumbens) in the final models is also worth more discussion. Perhaps it has been overly emphasized in the past, but there are strong results from other studies pointing to the importance of reward circuitry.

The design involved a nice time series for collecting behavioral and neural data at four time points: 0 hr (mating), 2.5 hr (mating huddling, investigation). 6 (early unstable bond) and 22 hours after mating.

Please discuss the consequences of creating the behavioral data for pair bond formation by subtracting same-sex pairs interactions from the opposite-sex interactions. What sources of information are removed by using this approach?

Time 0 is when the barrier is removed after a two-hour exposure. Please speculate on what is going on during the two-hour exposure. Time zero is potentially more than the time of mating. Is it possible that aggression is being decreased during this time point that represents mating? Could it also be a measure of the outcome of an initial compatibility assessment by the male and female?

Reviewer #2 (Public Review):

In this manuscript, the authors generate an annotated brain atlas for the prairie vole, which is a widely studied organism. This species has a suite of social behaviors that are difficult or impossible to study in conventional rodents, and has attracted a large community of researchers. The atlas is impressive and will be a fantastic resource. The authors use this atlas to examine brain-wide c-fos expression in prairie voles that were paired with same-sex or opposite-sex voles across multiple time points. In some sense, the design resembles PET studies done in primates that take whole brain scans after an important behavioral experience. The authors observed increased c-fos expression across a network of brain regions that largely corresponds with the previous literature. The study design captured several novel observations including that c-fos expression in some regions correlates strongly between males and females during pair bond formation and mating, suggesting synchrony in neural activity. An important caveat to this study not mentioned by the authors is that c-fos provides a snapshot of neural activity and that important populations of neurons could be active and not express c-fos. Thus observed correlations are likely to be robust, but the absence of differences (in say accumbens) may just reflect the limits of c-fos estimation of neural activity. Similarly, highly coordinated neural activity between males and females might still be driven by different mechanisms if different cell types were activated within a specific region. Nonetheless, the creation of this resource and its use in a well-designed study is an important accomplishment.

Reviewer #3 (Public Review):

In this manuscript, Gustison et al., describe the development of an automated whole-brain mapping pipeline, including the first 3D histological atlas of the prairie vole, and then use that pipeline to quantify Fos immunohistochemistry as a measure of neural activity during mating and pair bonding in male and female prairie voles. Prairie voles have become a useful animal model for examining the neural bases of social bonding due to their socially monogamous mating strategy. Prior studies have focused on identifying the role of a few neuromodulators (oxytocin, vasopressin, dopamine) acting in a limited number of brain regions. The authors use this unbiased approach to determine which areas become activated during mating, cohabitation, and pair bonding in both sexes to identify 68 brain regions clustered in seven brain-wide neuronal circuits that are activated over the course of pair bonding. This is an important study because i) it generates a valuable tool and analysis pipeline for other investigators in the prairie vole research community and ii) it highlights the potential involvement of many brain regions in regulating sexual behavior, social engagement, and pair bonding that have not been previously investigated.

Strengths of the study include the unbiased assessment of neural activity using the automated whole brain activity mapped onto the 3D histological atlas. The design of the behavioral aspect of the study is also a strength. Brains were collected at baseline and 2.5, 6 and 22 hrs after cohabitation with either a sibling or opposite-sex partner. These times were strategically chosen to correspond to milestones in pair bond development. Behavior was also quantified during epochs over the 22 hr period providing useful information on the progression of behaviors (e.g. mating) during pair bonding and relating Fos activation to specific behaviors (e.g. sex vs bonding). The sibling co-housed group provided an important control, enabling the identification of areas specifically activated by sex and bond formation. The analyses of the data were rigorous, resulting in convincing conclusions. While there was nothing particularly surprising in terms of the structures that were identified to be active during the mating and cohabitation, the statistical analysis revealed interesting relationships in terms of interactions of the various clusters, and also some level of synchrony in brain activation between partners. Furthermore, ejaculation was found to be the strongest predictor of Fos activation in both males and females. The sex differences identified in the study were subtle and less than the authors expected, which is interesting.

While the study provides a potentially useful tool and approach that may be of general use to the prairie vole community and identifies in an anatomically precise manner areas that may be important for mating or pair bond formation, there are some weaknesses as well. The study is largely descriptive. It is impossible to determine whether the activated areas are simply involved in sex or in the pair bond process itself. In other words, the authors did not use the Fos data to inform functional testing of circuits in pair bonding or mating behaviors. However, that is likely beyond the scope of this paper in which the goal was more to describe the automated, unbiased approach. This weakness is offset by the value of the comprehensive and detailed analysis of the Fos activation data providing temporal and precise anatomical relationships between brain clusters and in relation to behavior. The manuscript concludes with some speculative interpretations of the data, but these speculations may be valuable for guiding future investigations.