Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Feb 21:12:RP87029.
doi: 10.7554/eLife.87029.

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

Morgan L Gustison  1   2 , Rodrigo Muñoz-Castañeda  3   4 , Pavel Osten  3 , Steven M Phelps  1   5
Affiliations

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

Morgan L Gustison et al. Elife. .

Abstract

Sexual bonds are central to the social lives of many species, including humans, and monogamous prairie voles have become the predominant model for investigating such attachments. We developed an automated whole-brain mapping pipeline to identify brain circuits underlying pair-bonding behavior. We identified bonding-related c-Fos induction in 68 brain regions clustered in seven major brain-wide neuronal circuits. These circuits include known regulators of bonding, such as the bed nucleus of the stria terminalis, paraventricular hypothalamus, ventral pallidum, and prefrontal cortex. They also include brain regions previously unknown to shape bonding, such as ventromedial hypothalamus, medial preoptic area, and the medial amygdala, but that play essential roles in bonding-relevant processes, such as sexual behavior, social reward, and territorial aggression. Contrary to some hypotheses, we found that circuits active during mating and bonding were largely sexually monomorphic. Moreover, c-Fos induction across regions was strikingly consistent between members of a pair, with activity best predicted by rates of ejaculation. A novel cluster of regions centered in the amygdala remained coordinated after bonds had formed, suggesting novel substrates for bond maintenance. Our tools and results provide an unprecedented resource for elucidating the networks that translate sexual experience into an enduring bond.

Keywords: Microtus ochrogaster; c-Fos; iDISCO; immediate early gene; neuroscience; pair bond; prairie vole.

PubMed Disclaimer

Conflict of interest statement

MG, RM, SP No competing interests declared, PO P.O. is a co-founder, shareholder and Director at Certerra, Inc and co-founder, shareholder and President at Certego Therapeutics, Inc

Figures

Figure 1.
Figure 1.. Prairie vole reference brain and atlas generation for automatic c-Fos analysis.
(A) Generation of prairie vole (upper row) and mouse (lower row) reference brains was done with light-sheet fluorescence microscopy (LSFM) imaging. Shown is the top view of both prairie vole and mouse brains after perfusion. Also shown are cross-sections of coronal and horizontal views of prairie vole and mouse reference brains built after the co-registration of nearly 200 brains per species (approximately 190 voles and 150 mice). The prairie vole whole brain was 1.39 times larger than the mouse brain (see Figure 1—figure supplement 1 and Supplementary file 1). Both prairie vole and mouse brains underwent 3D renderization. Also shown are sagittal views of prairie vole and mouse whole-brain fluorescent Nissl (NeuroTrace) staining imaged with LSFM. The scale bar represents a distance of 1,000 μm. (B) Whole-brain staining of both prairie vole (upper row) and mouse (lower row) brains were registered to the reference brain. In red, boundaries of registered mouse reference atlas are plotted onto both prairie vole and mouse reference brains. NeuroTrace, tyrosine hydroxylase (TH), and somatostatin (SST) were registered and overlaid onto the atlases for validation. (C) Coronal sections of the resulting prairie vole atlas after manual validation. Scale bars represent distances of 1,000 μm and 200 μm for whole section and close-up views, respectively. (D) Overview of LSFM prairie vole c-Fos analysis pipeline. Shown in the left panel is a sagittal section of prairie vole c-Fos immunolabeling imaged with LSFM (scale bar represents a distance of 1,000 μm). Shown in the center panel are detailed views of two brain locations immunolabeled with c-Fos and overlaid with the resulting segmentation. Shown in the right panel is an area-based analysis of c-Fos+ cells. All c-Fos+ cell centroids are registered to the prairie vole reference brain and analyzed using the new prairie vole reference atlas. For each brain, a voxel representation is generated of all c-Fos+ cells in the same prairie vole reference space and overlaid with the reference atlas.
Figure 1—figure supplement 1.
Figure 1—figure supplement 1.. Validation of the prairie vole reference atlas.
Mouse coronal sections (left column) are composed of the mouse reference brain overlaid with atlas boundaries in red on the left and, in the same sections, of tyrosine hydroxylase (TH) immunolabeling on the right. Prairie vole coronal sections are in the right three columns. In the first vole column, coronal brain sections are of the prairie vole reference brain and atlas boundaries in red on the left with TH immunolabeling on the right. In the second vole column, coronal sections are of the prairie vole reference brain and c-Fos+ mean voxelization overlay with atlas boundaries in red. In the third vole column, coronal sections of prairie vole NeuroTrace staining are registered onto the prairie vole reference brain. Scale bars represent a distance of 1,000 μm.
Figure 2.
Figure 2.. Study design and development of the prairie vole pair bond.
(A) Schematic of the experiment design is shown, where social behaviors (during 1 hr observations, black blocks) and immediate early gene (IEG) expression patterns are compared between mate pairs and siblings across time. (B) Continuous automated tracking is shown for active social interactions (i.e., an index of social investigation and mating) in mate pairs (red lines) and sibling pairs (blue lines). (C) Time courses are shown for specific social behaviors during 1 hr observation windows that correspond to IEG expression, with each row representing one pair. Social behaviors are overlaid onto self-grooming (grey), and ultrasonic vocalizations (USVs, short dark red ticks) are overlaid onto all behaviors. Other behaviors shown include anogenital investigation (light blue), close follows (dark blue), mount (very light red), intromissions (light red), ejaculation (red), huddle (light green), and allogroom (dark green). (D) Plots show group differences (mean ± standard deviation [SD]) in individual activity level (velocity) and movement relative to the partner (net move, positive values indicate movement toward the partner). (E) Group differences (mean ± SD) are shown for vocal behavior (ultrasonic vocalizations [USVs]), proximity seeking (approach), mating (mount), and side-by-side contact (huddle). For (D, E), mate pairs are in red and sibling pairs in blue for dyadic-level behaviors (USVs, mount, and huddle). Females are a lighter hue and males a darker hue for behaviors measured on an individual level (velocity, netmove, approach). T-tests were used to compare mate animals (n = 94) or mate dyads (USVs: n = 48; mount and huddle: n = 47) with sibling animals (n = 96) or sibling dyads (n = 48). Paired t-tests were used to compare female mates (n = 47) with male mates (n = 47). For behaviors with group-level effects, follow-up t-tests were performed for each timepoint (n = 11-12 animals/group or n = 22-24 dyads/group; significance represented by * p<0.05, ** p<0.01, *** p<0.001). Descriptions of vole behaviors are listed in Supplementary file 3.
Figure 2—figure supplement 1.
Figure 2—figure supplement 1.. Automated tracking of social behavior states.
(A) Long-term tracking of social states was informed by automated measures of the largest body area, pair activity, and pair distance. On the left, body area is plotted against dyad distance from an exemplar 24 hr mate pair video (10% randomly selected frames from 12 hr of white light). On the right, the density curve for body area reveals a basic threshold for when animals are separated or in physical contact. (B) Traces of body area, video activity, and pair distance are plotted for the first hour of interaction in the exemplar mate pair, along with automated assignments of behavioral states. (C) Cumulative time spent in inactive social contact is shown for up to 22 hr of cohabitation in mate pairs (red lines) and sibling pairs (blue lines).
Figure 2—figure supplement 2.
Figure 2—figure supplement 2.. Time course of social behaviors during pairing.
Group differences (mean ± standard deviation [SD]) are shown for appetitive behaviors including nose-to-nose touching, anogenital investigation, and close follows. Group differences (mean ± SD) are also shown for proximity and grooming behaviors including pair distance, allogrooming, and selfgrooming. Mate pairs are in red and sibling pairs are in blue for dyadic-level behaviors (nose touch and pair distance). For individual-level behaviors (investigate, follow, allogroom, and selfgroom), females are shown in lighter hues, and males in darker hues. T-tests were used to compare mate animals (n = 94) or mate dyads (n = 47) with sibling animals (n = 96) or sibling dyads (n = 48). Paired t-tests were used to compare female mates (n = 47) with male mates (n = 47). For behaviors with main effects, follow-up t-tests were performed for each timepoint (n = 11-12 animals/group or n = 22-24 dyads/group; significance represented by * p<0.05, ** p<0.01). Descriptions of vole behaviors are listed in Supplementary file 3.
Figure 2—figure supplement 3.
Figure 2—figure supplement 3.. Associations between behavioral states and types of social interaction.
Hierarchical clustering of behavioral measures from Pearson correlations is used to group behavioral states and social interactions into three main clusters (n = 190 animals). These clusters involve close contact (e.g., huddling and allogrooming), mating (e.g., mounts and vocalizations), and appetitive behavior (e.g., approaches and follows). On the right, Pearson correlations are shown among behaviors in subsets based on partner type, sex, and timepoint (n = 11-12 animals/group). Warm and cool colors indicate positive and negative correlation coefficients, respectively. White indicates behaviors with no variation, meaning that coefficients were not calculated (e.g., siblings did not exhibit mating behaviors). Descriptions of vole behaviors are listed in Supplementary file 3.
Figure 3.
Figure 3.. A brain-wide functional network is active during pair bond formation.
(A) Map of bonding-associated voxels is shown in coronal sections (rostral to caudal) spanning the vole brain. Brightness indicates significance levels for comparisons of hypothesized and null generalized linear models (GLMs) to predict c-Fos+ cell counts across 189 vole brain samples. For each voxel, an analysis of variance (ANOVA) test was used to compare null and hypothesized GLMs. False discovery rate (FDR) q-values were computed from the ANOVA results to account for multiple comparisons across voxels (alpha threshold of q = 0.1). Warm colored voxels have higher mean c-Fos+ cell counts in mates (n = 94) compared to siblings (n = 95) (GLM test statistic for partner type > 0). Cool colored voxels have lower mean c-Fos+ cell counts in mates compared to siblings (GLM test statistic for partner type < 0). (B) Identification of significant and mutually exclusive regions of interest (ROIs, n = 68 chosen and 824 total ROIs) is shown sorted by anatomical division and F-statistic. For each ROI, an ANOVA test was used to compare null and hypothesized GLMs. The significance levels of ANOVAs were determined by comparing the observed F-statistic to its null distribution from shuffled data (10,000 permutations). False discovery rate (FDR) q-values were computed to adjust for multiple comparisons acros ROIs (alpha threshold of q = 0.1). Colored symbols for ROIs match their cluster group assignments in (C) and (D). (C) Hierarchical clustering of chosen ROIs (n = 68, permutation tests with FDR q < 0.1) is shown alongside pairwise Pearson correlations of c-Fos+ cell counts. Full ROI names are listed in Supplementary file 4. (D) Multi-dimensional scaling (MDS) coordinate space of the correlations is shown between chosen ROIs (n = 68). The most significant ROIs per cluster are labeled with the symbol size scaled by the F-statistic. Darkness and thickness of connecting lines reflect Pearson correlation coefficients. (E) Time course trajectories of total c-Fos+ cell counts are shown for each cluster. Counts per cluster are scaled across samples with means taken for each experiment group (n = 11-12 animals/group). Red lines represent mates and blue lines represent siblings (females – lighter, males – darker). In (C, D, E), each cluster group is given a label to summarize the most significant ROIs within it. Cluster group labels include ‘BST’ (bed nucleus of the stria terminalus) in dark purple, ‘AOB’ (accessory olfactory bulb) in light purple, ‘PFC’ (prefrontal cortex) in green, ‘POA-VMH’ (preoptic area, ventromedial hypothalamus) in dark blue, ‘PVH’ (paraventricular hypothalamus) in light blue, ‘Thal’ (thalamus) in red, ’Aud’ (auditory cortex) in light orange, and ‘Amyg’ (amygdala) in dark orange.
Figure 3—figure supplement 1.
Figure 3—figure supplement 1.. Brain-wide patterns of immediate early gene activation during pairing.
(A) Coronal cross-sections (rostral to caudal) from female (top) and male (bottom) mating pairs are shown for the 2.5 hr timepoint group, with brightness corresponding to the mean voxel c-Fos+ cell counts (n = 12 female and 12 male mates). (B) A representative coronal slice that includes the posterior part of the bed nucleus of the stria terminalis is shown for mate pairs and siblings and separated by timepoint and sex. Brightness corresponds to mean voxel c-Fos+ cell counts per group (n = 11-12 animals/group).
Figure 3—figure supplement 2.
Figure 3—figure supplement 2.. Patterns of immediate early gene expression in brain region clusters.
Counts of c-Fos+ cells are shown for regions of interest (ROIs) associated with pair-bond development (n = 68), organized by hierarchical cluster. For each cluster, the total counts are shown on the right and the most significant ROI (highest F-statistic) is shown on the right. For each ROI, an analysis of variance (ANOVA) test was used to compare null and hypothesized generalized linear models (GLMs). The significance levels of the ANOVAs were determined by comparing the observed F-statistic to its null distribution from shuffled data (10,000 permutations). False discovery rate (FDR) q-values were computed to adjust for multiple comparisons across ROIs (alpha threshold of q = 0.1). Counts are summarized by timepoint, partner type, and sex (mean ± standard deviation [SD], n = 11-12 animals/group). Mate pairs are in red and sibling pairs in blue (females in lighter hues, males in darker hues). Cluster group labels include ‘BST’ (bed nucleus of the stria terminalus), ‘POA-VMH’ (preoptic area, ventromedial hypothalamus), ‘PVH’ (paraventricular hypothalamus), ‘AOB’ (accessory olfactory bulb), ‘Thal’ (thalamus), ‘PFC’ (prefrontal cortex), ‘Amyg’ (amygdala), and ’Aud’ (auditory cortex). Representative ROIs include PD (posterodorsal preoptic nucleus), MPO (medial preoptic area), PVpo (periventricular hypothalamic nucleus, preoptic part), y (nucleus y), BAC (bed nucleus of the anterior commisure), PL5 (prelimbic area layer 5), POR (superior olivary complex, periolivary region), and AUDv6b (ventral auditory area layer 6b).
Figure 3—figure supplement 3.
Figure 3—figure supplement 3.. Multi-dimensional structure of brain-wide correlation patterns.
Representations of the first three dimensions of a multi-dimensional scaling (MDS) coordinate space based on Pearson correlations between c-Fos+ cell counts in brain regions of interest (ROIs) associated with bonding (n = 68 ROIs from 189 vole brain samples). Each symbol represents a ROI and is colored based on cluster assignment. Darkness and thickness of connecting lines reflect correlation coefficients. Cluster group labels include ‘BST’ (bed nucleus of the stria terminalus) in dark purple, ‘AOB’ (accessory olfactory bulb) in light purple, ‘PFC’ (prefrontal cortex) in green, ‘POA-VMH’ (preoptic area, ventromedial hypothalamus) in dark blue, ‘PVH’ (paraventricular hypothalamus) in light blue, ‘Thal’ (thalamus) in red, ’Aud’ (auditory cortex) in light orange, and ‘Amyg’ (amygdala) in dark orange. Full ROI names are listed in Supplementary file 4.
Figure 3—figure supplement 4.
Figure 3—figure supplement 4.. Anatomical connectivity in brain regions associated with pairing.
(A) This heatmap shows normalized connection densities (Knox et al., 2019) from projecting to ipsilateral target regions of interest (ROIs) in the mouse brain. These ROIs are the same as, or larger divisions that contain, the ROIs found to be associated with pair-bond development in our analysis. (B) The top histogram shows results of a permutation test to assess whether hierarchical clusters of chosen ROIs (n = 68 from 189 vole brain samples) in our analysis mirror underlying anatomical connections. Observed connectivity reflects the overall average of cluster means, which is compared to averages from 10,000 iterations of shuffled data (i.e., target regions shuffled for each projecting region). The bottom histograms show the results of permutation tests for mean connectivity within each ROI cluster. Cluster group labels include ‘BST’ (bed nucleus of the stria terminalus) in dark purple, ‘AOB’ (accessory olfactory bulb) in light purple, ‘PFC’ (prefrontal cortex) in green, ‘POA-VMH’ (preoptic area, ventromedial hypothalamus) in dark blue, ‘PVH’ (paraventricular hypothalamus) in light blue, ‘Thal’ (thalamus) in red, ’Aud’ (auditory cortex) in light orange, and ‘Amyg’ (amygdala) in dark orange. Full ROI names are listed in Supplementary file 4.
Figure 4.
Figure 4.. Bed nucleus of the stria terminalis (BST) emerges as a central hub in the bonding network that is associated with mating success in both sexes.
(A) The first dimension of canonical correlation (CC) scores is compared across experiment groups (mean ± standard deviation [SD], n = 11-12 animals/group). (B) Heatmaps represent correlation coefficients among CC scores, region of interest (ROI) c-Fos+ cell counts, and behavior measures. The full dataset is on top (n = 189 animals for correlations involving ultrasonic vocalizations [USVs], n = 187 for all other behaviors), and the bottom two correlograms are for female (n = 23) and male (n = 24) mates. In (B, E, F), warm and cool colors represent positive and negative coefficients, respectively. Arrows mark the ROI (principle nucleus of the BST [BSTpr]) and behavior (ejaculation) with the strongest correlations to CC1. (C) BSTpr activation is compared across groups (n = 11-12 animals/group) with a GLM comparison (ANOVA) followed by a permutation test (10,000 permutations) and false discovery rate (FDR) correction for multiple tests across ROIs. (D) Successful mating events are shown across timepoints (mean ± SD, n = 22-24 dyads/group). (E) Similarity (i.e., Pearson correlation) of bonding network activation is shown in female–male pairs at the 0 hr (n = 11), 2.5 and 6 hr (n = 23) and 22 hr (n = 12) timepoints. (F) Similarity in bonding network activation is shown in female–male pairs (n = 23), using partial Pearson correlations to control for ejaculation rates. (G) Female–male pair similarity is shown for CC1 scores (n = 24 pairs) and BSTpr activation (n = 23 pairs) during bond formation. (H) Pearson correlations are used to associate ejaculation rates with BSTpr activation for female mates (n = 23) and male mates (n = 24). In (A, C, D), mate pairs are in red and sibling pairs in blue (females in lighter hues, males in darker hues). In (B, E, F), cluster group labels include ‘BST’ (bed nucleus of the stria terminalus) in dark purple, ‘AOB’ (accessory olfactory bulb) in light purple, ‘PFC’ (prefrontal cortex) in green, ‘POA-VMH’ (preoptic area, ventromedial hypothalamus) in dark blue, ‘PVH’ (paraventricular hypothalamus) in light blue, ‘Thal’ (thalamus) in red, ’Aud’ (auditory cortex) in light orange, and ‘Amyg’ (amygdala) in dark orange. ROI names and order correspond to Figure 3C. Full ROI names are listed in Supplementary file 4.
Figure 4—figure supplement 1.
Figure 4—figure supplement 1.. Dimensions of cross-covariance in immediate early gene expression and social behavior.
(A) A schematic of the canonical correlation analysis (CCA) method is shown, where two sets of variables (x and y) are combined linearly to extract latent variates with the strongest correlations (CC1 stronger than CC2, and so on). This analysis outputs x and y scores for each latent variate for each animal subject. In this dataset, x scores represent linear combinations of behavioral measures, and y scores represent linear combinations of region of interest (ROI) c-Fos+ cell counts. (B) On the left, the relationship between x and y scores is shown for all study animals for the first three variates (n = 191 animals). On the right, the means of x and y scores for the first three variates are compared across partner type, sex, and timepoint (n = 11-12 animals/group; boxplots = median, 25–75% quartile, and 95% confidence intervals). Mate pairs are in red and sibling pairs in blue (females in lighter hues, males in darker hues).
Figure 4—figure supplement 2.
Figure 4—figure supplement 2.. Intra-pair similarity in immediate early gene expression in pairing-associated brain regions.
Heatmaps represent Pearson correlation coefficients, with brain regions on the y-axis (grouped into hierarchical clusters, see Figure 3 for region of interest [ROI] and cluster labels) and behavioral outcomes (during 1 hr observation windows) on the x-axis. Correlation heatmaps are split by partner type, sex, and timepoint (n = 11-12 animals/group). Warm and cool colors indicate positive and negative correlation coefficients, respectively. White indicates behaviors with no variation, meaning that coefficients were not computed (e.g., siblings did not exhibit mating behaviors). ROIs are grouped by hierarchical clusters (see Figure 3). Cluster group labels include ‘BST’ (bed nucleus of the stria terminalus) in dark purple, ‘AOB’ (accessory olfactory bulb) in light purple, ‘PFC’ (prefrontal cortex) in green, ‘POA-VMH’ (preoptic area, ventromedial hypothalamus) in dark blue, ‘PVH’ (paraventricular hypothalamus) in light blue, ‘Thal’ (thalamus) in red, ’Aud’ (auditory cortex) in light orange, and ‘Amyg’ (amygdala) in dark orange. ROI names and order correspond to Figure 3C. Full ROI names are listed in Supplementary file 4.
Figure 4—figure supplement 3.
Figure 4—figure supplement 3.. Patterns of association between immediate early gene expression and behavior.
On the top, heatmaps represent Pearson correlation coefficients between selected brain regions of interest (ROIs) in mating pairs (n = 11-12 pairs/timepoint), with female data on the y-axis and male data on the x-axis. On the bottom, heatmaps represent Pearson correlation coefficients between selected ROIs in sibling dyads (n = 11-12 dyads/timepoint), where sibling A data are from animals placed in the left side of chamber during acclimation, and sibling B data are from animals placed on the right side. ROIs are grouped by hierarchical clusters (see Figure 3). Warm and cool colors indicate positive and negative correlation coefficients, respectively. Cluster group labels include ‘BST’ (bed nucleus of the stria terminalus) in dark purple, ‘AOB’ (accessory olfactory bulb) in light purple, ‘PFC’ (prefrontal cortex) in green, ‘POA-VMH’ (preoptic area, ventromedial hypothalamus) in dark blue, ‘PVH’ (paraventricular hypothalamus) in light blue, ‘Thal’ (thalamus) in red, ’Aud’ (auditory cortex) in light orange, and ‘Amyg’ (amygdala) in dark orange. ROI names and order correspond to Figure 3C. Full ROI names are listed in Supplementary file 4.
Figure 5.
Figure 5.. Working model for neural systems that shape stages of pair-bond development.
(A) A schematic is shown of the network of regions identified in our study, overlaid with regions involved in rodent mating behavior (Pfaus and Heeb, 1997; Veening et al., 2014; Veening and Coolen, 2014). This network is proposed to be involved during the early stages of bond formation (e.g., 2.5 hr timepoint). (B) A schematic is shown of the network of regions identified in our study, overlaid with regions involved in prairie vole pair-bonding behavior (Walum and Young, 2018; Young et al., 2011). This network is proposed to be involved with the middle stages of bond formation when animals are engaged in prolonged mating and affiliative interactions (e.g., 6 hr timepoint). (C) A schematic is shown of the network of regions that are correlated between female and male mates at the 22 hr timepoint in this study. These regions are identified from pairs of regions in which both sexes show high inter-individual similarity (Pearson correlation, r > 0.75). This network is proposed to be involved with the recognition and convergence of behavioral state in bonding partners. The network schematics in (A–C) are adapted from the multi-dimensional scaling of correlations between region activity (Figure 3D). Light gray points represent regions not included in the proposed network. Line thickness and darkness between colored regions represent Pearson correlation r values between the connected regions. Cluster group labels include ‘BST’ (bed nucleus of the stria terminalus) in dark purple, ‘AOB’ (accessory olfactory bulb) in light purple, ‘PFC’ (prefrontal cortex) in green, ‘POA-VMH’ (preoptic area, ventromedial hypothalamus) in dark blue, ‘PVH’ (paraventricular hypothalamus) in light blue, ‘Thal’ (thalamus) in red, ’Aud’ (auditory cortex) in light orange, and ‘Amyg’ (amygdala) in dark orange. Full ROI names are listed in Supplementary file 4.
See this image and copyright information in PMC

Update of

Similar articles

See all similar articles

References

    1. Amadei EA, Johnson ZV, Jun Kwon Y, Shpiner AC, Saravanan V, Mays WD, Ryan SJ, Walum H, Rainnie DG, Young LJ, Liu RC. Dynamic corticostriatal activity biases social bonding in monogamous female prairie voles. Nature. 2017;546:297–301. doi: 10.1038/nature22381. - DOI - PMC - PubMed
    1. Aragona BJ, Liu Y, Yu YJ, Curtis JT, Detwiler JM, Insel TR, Wang Z. Nucleus accumbens dopamine differentially mediates the formation and maintenance of monogamous pair bonds. Nature Neuroscience. 2006;9:133–139. doi: 10.1038/nn1613. - DOI - PubMed
    1. Baumeister RF, Leary MR. The need to belong: desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin. 1995;117:497–529. - PubMed
    1. Bayless DW, Yang T, Mason MM, Susanto AAT, Lobdell A, Shah NM. Limbic neurons shape sex recognition and social behavior in sexually naive males. Cell. 2019;176:1190–1205. doi: 10.1016/j.cell.2018.12.041. - DOI - PMC - PubMed
    1. Bendesky A, Kwon YM, Lassance JM, Lewarch CL, Yao S, Peterson BK, He MX, Dulac C, Hoekstra HE. The genetic basis of parental care evolution in monogamous mice. Nature. 2017;544:434–439. doi: 10.1038/nature22074. - DOI - PMC - PubMed

Substances