<i>Agouti</i> integrates environmental cues to regulate paternal behaviour

Nature News · Feb 18, 2026 · Collected from RSS

Full Article

MainActive parental care is crucial for the survival and development of altricial offspring, with maternal care forming the evolutionary foundation of caregiving across all mammals. In all mammalian species, mothers provide essential resources such as nutrition, thermoregulation and protection4. In a substantially smaller fraction of mammals (3–5%), including humans1,5, fathers also contribute to offspring care, which may confer unique developmental and evolutionary advantages. Although the neural circuits underpinning maternal care have been extensively studied and are conserved across mammals, our understanding of the causal neurobiological and molecular mechanisms that give rise to naturally occurring, active care in males is still in its infancy1.Examination of neuroendocrine and peptidergic pathways of known importance in lactating mothers, infanticidal and paternal behaviour in standard laboratory mice (Mus musculus) and rats (Rattus norvegicus), and genetic variation among deer mice and vole species have yielded essential insights into the paternal brain1,6,7. It is well-established that maternal care is mediated by the medial preoptic area of the hypothalamus (MPOA), a brain region that undergoes dynamic hormonal and epigenetic remodelling during pregnancy1. This remodelling reshapes neural circuits, suppressing pup-directed aggression while promoting approach and engaging downstream reward systems to reinforce caregiving8,9. By contrast, sexually naive male mice and rats are typically infanticidal in the wild10,11, although following mating, they temporarily reduce infanticidal behaviour and show paternal behaviours for several weeks12,13,14. Sexually naive male laboratory mice and rats can also be induced to show parental behaviours through sensitization by prolonged pup exposure13. In sensitized males, the MPOA has likewise been identified as a critical hub15,16,17, integrating upstream olfactory inputs16,17 and engaging downstream reward pathways18,19, suggesting that the fundamental neural circuitry for caregiving exists in males but remains inactive under normal conditions. Yet, there are undoubtedly neurobiological differences among species that are naturally caring before sexual experience, exemplified by work in cooperatively breeding voles1,20,21,22,23,24,25. Whereas these studies provide insight into how paternal behaviours can be elicited, they also underscore the need for new rodent models for understanding the neurobiological basis of naturally occurring variation in male caregiving, particularly within a single species.The African striped mouse (R. pumilio) is an attractive model for investigating the neurobiology of paternal care. Unlike traditional laboratory murine rodents, sexually naive African striped mice (hereafter, ‘striped mice’) naturally show both paternal and alloparental behaviours in the wild and in the laboratory, with non-genetic caregivers (alloparents) readily providing care to younger siblings and even unrelated infants2. Striped mice thrive in a diverse range of social environments, from communal group living to philopatry, territorial breeding and solitary wandering2,26,27,28. When male striped mice disperse from their natal nest, they assume one of a variety of social strategies according to prevailing environmental pressures such as population density and resource availability. These strategies include solitary roaming and the formation of bachelor groups (with kin and non-kin)26,29,30,31. This ecological flexibility extends to laboratory settings, in which they can be housed under conditions that mimic these social strategies. Notably, male striped mice do not require previous reproductive experience or sensitization to show parental care28, and their parental phenotypes show natural intraspecies variation. This marked variation within a single species, in which genetic factors are controlled, offers a unique opportunity to dissect the molecular and neurobiological mechanisms underlying experience-dependent individual differences in paternal care. Here, leveraging the striped mouse as a model and using an integrative approach—including brain-wide mapping of cFos expression, single-nucleus RNA sequencing (snRNA-seq), viral-mediated gene manipulation in the brain and socio-environmental manipulations—we provide new insight into the neural mechanisms driving naturally occurring paternal care.Striped mouse males are primed to careWhereas previous field and laboratory studies have shown that sexually naive male striped mice (Fig. 1a) show paternal behaviour2,32,33, the full extent and quality of this care as well as how it compares with that of genetic fathers (that is, sires) and mothers (that is, dams) in the laboratory setting has not been fully characterized. To this end, we compared the parental behaviour of laboratory-reared sexually naive males with that of sires and dams in a pup interaction test34. The temporal dynamics of behaviour were highly stereotyped across sex and reproductive experience (Supplementary Video 1), with nearly all mice initially crossing the enclosure to inspect (that is, sniff) pup stimuli with a subsequent transition into either caregiving or infanticidal behaviours. Care behaviours included licking and grooming, huddling contact and lateral contact (Extended Data Fig. 1 and Supplementary Table 1).Fig. 1: Allopaternal care in African striped mice.a, Photograph of an African striped mouse. b, Schematic of experimental design and behavioural testing. c, Survival curve illustrating the proportion of males assuming a huddling posture over a novel pup stimulus, comparing SI (grey) and GH (black) rearing conditions (log-rank Mantel–Cox test, X2 = 16.86, d.f. = 1, P < 0.0001). d, Cumulative affiliative contact time (sum of huddling, licking or grooming and adjacent contact time) with pups across rearing conditions (unpaired t-test, n = 46, t(45) = 4.063, P = 0.0002, Cohen’s d = 1.211, 95% confidence interval (180.2, 534.3)). e, Proportion of infanticide in males from different rearing conditions (log-rank Mantel–Cox test, Χ2 = 5.521, d.f. = 1, P = 0.0188). f, Proportional representation of infanticidal, ambivalent or allopaternal responses towards pups (chi-square test on proportions: X2 = 54.16, d.f. = 2, P < 0.0001). g, Time spent in affiliative contact with other social stimuli, including male weanlings (P24; n = 26, t(25) = 4.453, P = 0.0002), adult females (n = 26, t(25) = 2.711, P = 0.0119) and unrelated adult males (n = 26, t(25) = 0.7597, P = 0.4545). h, Time spent in the centre of an open field (mixed effects model: n = 46, t(45.00) = −1.486, P = 0.144). i, Time in the light portion of a light/dark box (n = 51, t(50) = 2.009, P = 0.0500). n represents the number of independent mice. Data in d,g,h,i are presented as mean ± s.e.m. All t-tests were two-sided and independent. ALLO, allopaternal; AMBI, ambivalent; INFA, infanticidal. Photograph in a by C. Todd Reichart, Department of Molecular Biology, Princeton University.Full size imageCollectively, males showed a wide range of phenotypes, from ‘dyspaternal’ (pup ambivalence or infanticide) to ‘allopaternal’. We defined ambivalence as neglect following an initial inspection of the pup stimulus (for example, Supplementary Video 2), whereas allopaternal males showed licking, grooming and huddling behaviours without previous reproductive experience at levels comparable to sires and dams (Extended Data Fig. 1, Supplementary Data Table 2 and Supplementary Video 3). Our results confirm that reproductive experience is not a prerequisite for paternal care in male striped mice.Social environment drives variation in male careIn the wild, striped mice show social flexibility, naturally forming either complex social groups or living in isolation, depending on environmental factors such as season and population density26,28,35, and previous studies have shown that social environment (for example, social isolation (SI) versus familial housing) can alter male care28. We leveraged this social flexibility to explicitly determine how social environment influences development of male parental care behaviours by rearing males either in SI or group housing (GH) with three to four age-matched males from weaning until sexual maturity (greater than or equal to PND80) (Fig. 1b). This developmental period encompasses the window in which many striped mouse males would disperse from their natal nest to roam in isolation, find their own territory or join a bachelor group27,29,30,31,36. In the pup interaction test, SI males engaged in significantly higher levels of all individual paternal behaviours (that is, huddling, licking or grooming and other pup contact) than GH males (Fig. 1c–e). Social environment also shaped overall paternal phenotypes: 21% of GH males were allopaternal, whereas 29% were infanticidal, with the remaining ambivalent (Fig. 1f). By contrast, 65% of SI males were allopaternal and none showed infanticide (Fig. 1f).To determine whether differences in allopaternal care could be attributed to baseline social, exploratory or anxiety-like behaviours driven by rearing environment, we assessed males in ten comprehensive behavioural tests (Fig. 1g–i and Extended Data Fig. 2). Together, these further tests suggest that isolation increases pro-reproductive behaviours, and these changes cannot be fully explained by differences in general sociality, anxiety, exploration or novelty preference.The MPOA is a hub of male parentingTo identify brain regions implicated in paternal behaviour in striped mice and determine whether neural activity in these regions is associated with individual differences, we exposed sexually naive GH and SI striped mouse males to either pups (n = 11 per group) or an empty cage (n = 10 per group) for 60 minutes. We then quantified brain-wide cFos expression as a marker of neural activity (Fig. 2a). Regardless of rearing condition, pup-exposed mice showed greater cFos than controls in