On a Wild Mouse Chase to Understand Parenting, Love, and Fear todayheadline

One of Aesop’s fables tells of two cousin mice, one from the country and the other from the city. Though related, their habits and preferences are remarkably different. The country mouse enjoys simple meals of beans, cheese, and bread in the peace of his rural home. Meanwhile, the city mouse feasts on ale, jellies, and cakes, but must constantly dodge imposing and bothersome mastiffs while dining.

Much like the protagonists of The Town Mouse and the Country Mouse, a real-life genus of agile, closely related North American wild mice displays strikingly diverse behaviors. Mice of the genus Peromyscus range from monogamous partners and devoted parents to promiscuous and uninvolved ones. Some build intricate burrows, while others keep them simple. When faced with threats, some rush to flee, while others delay this response.

Thus, unlike the popular lab mouse Mus musculus, which is often genetically modified to exhibit specific behavioral traits, Peromyscus species showcase a rich natural variation that has helped researchers uncover details of the genetic and neural basis of behavior. For instance, over a decade ago, scientists discovered that the preference for building complex versus small, simple burrows in two Peromyscus species was partly genetically determined and linked to specific genomic regions.¹ Now, research on this wild mouse family is shedding light on how love, care for offspring, and responses to threats are encoded in the animals’ DNA and brains.

A Distinctive, Though not New, Model to Study Variation

The remarkable diversity and wide distribution of Peromyscus mice across North America, along with their many local variations, have captivated naturalists for over a century. Now, scientists are leveraging advances in molecular techniques to peer into the neurobiology underpinning the unique behaviors of these nontraditional model species. 

“The beauty is that we can tap into this diversity in behavior,” said Felix Baier, a biologist at the Max Planck Institute for Brain Research. 

Unlike inbred lab mice commonly used in research, most scientists studying Peromyscus either capture them from the wild or use mice derived from lab colonies that were initially started with wild-caught individuals and maintained as outbred strains. “We can have this wild or wild-derived animal in a controlled lab environment, and then apply all the rigorous testing to understand the mechanisms that give rise to the diversity in behavior,” said Baier. 

Furthermore, since most Peromyscus species diverged only a few million years ago, many of these lineages can interbreed, producing hybrids. Studying these mixed offspring facilitates genetic analyses that can pinpoint regions in the genome linked to behavioral traits specific to one of the parental strains. 

Peromyscus’ natural variation and interbreeding are quite unique in terms of model organisms out there, said Baier.

A Tale of Different Parenting Styles

Among Peromyscus species are two monogamous ones, the oldfield mouse P. polionotus and the California mouse P. californicus. This is a rarity since only about nine percent of mammalian species display such behavior.² Consistent with the strong pair bonding in those two species, both mothers and fathers actively participate in raising their young, licking, grooming, and building a nest.

While doing a postdoctoral fellowship in the lab of Hopi Hoekstra at Harvard University, geneticist and neuroscientist Andrés Bendesky, who now leads his own lab at Columbia University, ran comparison studies between one of these monogamous species, P. polionotus, and a promiscuous species, the eastern deer mouse P. maniculatus. In contrast to P. polionotus, eastern deer mouse fathers provide minimal parental care, leaving most responsibilities to their female partners. Even mothers in this species show lower levels of involvement in most parenting behaviors compared to oldfield mouse mothers. 

Almost every behavior that researchers measure in these two species differs enormously, said Bendesky. “Yet they’re so closely related genetically that you can have a much more narrow space to look for the differences that matter,” he added. Both species diverged about 1.8 million years ago, making them as closely related as humans are to Neanderthals, he noted.

In 2017, Bendesky and his colleagues reported that parenting behavior in both species is largely inherited.³ The team removed pups from their biological parents and gave them to parents of opposite species. When the offspring grew up, their parenting strategies mirrored those of their biological parents, not their adoptive ones. For example, even though P. polionotus males were raised by adoptive fathers who rarely huddled or licked them, they still grew up to be nurturing fathers when it was their turn to care for their own offspring. 

By studying second-generation hybrids, the team identified genomic regions that contribute to these behavioral differences. They found, for instance, that variations in the gene coding for a precursor of vasopressin—a neuropeptide previously implicated in pair-bonding and maternal care—influenced the quality of nests built by parents, with higher expression associated with poorer nest quality.^4,5 Notably, the expression of this gene in the hypothalamus—a brain region key to the regulation of multiple social behaviors—is 2.8-fold higher in the promiscuous P. maniculatus than in the monogamous P. polionatus. 

To further explore the genetic and neural differences between these two species, Jenny Chen, a computational biologist in Hoekstra’s team, recently published a transcriptional cell atlas of the medial preoptic area (MPOA), a hypothalamic region that regulates mating and parenting behaviors.^6,7 The researchers found significant differences in the abundance of certain neuronal cell types across species. For instance, P. maniculatus expressed more vasopressin neurons, arguing once again for an important role of this neuropeptide in the species’ behavioral differences. 

Many scientists assume that, if evolutionary pressure favors an increase in vasopressin, it just needs to crank up gene expression, “But what we’re finding is that that’s not what nature decided to do,” said Chen. “What nature decided to do is to make [around two] times more vasopressin cells in the promiscuous species,” she added. The discovery raises several intriguing questions. Chen wondered whether increasing the number of neurons may enhance neural computational ability or if this shift reflects an evolutionary preference—perhaps it’s easier to evolve new cells rather than evolve higher gene expression levels, she hypothesized. 

When comparing neuronal gene expression in this region, Chen and her colleagues found that P. maniculatus males and females showed stronger differences between them than those exhibited by P. polionotus males and females. That is, sexual dimorphism in this molecular trait is reduced in the monogamous species. 

“There’s a strong prior literature on animals that have more pro-social behavior showing less sexual dimorphism just overall in their bodies, but also in these particular brain regions,” said Jessica Tollkuhn, a molecular biologist and neuroscientist at Cold Spring Harbor Laboratory who was not involved in this study. This new dataset supports ideas like this that have been floating around in the field for a long time and offers a tractable way to investigate them in the future, she said. 

Tollkuhn added, “There’s a lot of power in these comparative approaches because nature or natural selection has already done the work to create this biological variability and we can leverage this to understand the diversity of neural circuit function.” 

Yet, these parenting differences may not be solely rooted in variations within their brains. Over the years, while dissecting both species, Bendesky noticed that the monogamous P. polionotus had a massive adrenal gland compared to its promiscuous sister species. Intrigued, he pursued this observation now as leader of his own lab at Columbia University. His team found that P. polionotus evolved a unique adrenal cell type that expresses an enzyme that converts progesterone into 20α-hydroxyprogesterone.⁸ Moreover, injecting a dose of 20α-hydroxyprogesterone enhanced parental behaviors in oldfield mouse parents. This effect is likely driven, at least in part, by this metabolite’s conversion in the brain into a steroid that modulates inhibitory receptors involved in parental behavior.⁹ 

Interestingly, these adrenal gland differences appear to be exclusive to P. polionotus as the other monogamous species in the genus, P. californicus, lacks this cell type.

But, like the oldfield mouse, the California mouse has long served as a classic model for studying an engaged father—a rare opportunity given the scarcity of natural mammalian models exhibiting such behavior. Peeking into the brains of these devoted fathers has yielded valuable insights. For instance, researchers have reported changes in neurogenesis within the hippocampus of males as they interact with their offspring at different stages of the postpartum period.^10,11 Changes in neuronal proliferation and survival are well-documented in female mammals during the onset of motherhood, with some researchers linking these changes to maternal care and cognitive function, though their exact role remains uncertain.¹² Neural plasticity in P. californicus males is even less understood, and its potential connection to promoting paternal care is also unclear.¹³

Parenthood also fine-tunes sensory systems. Mammalian mothers, for example, undergo changes in hearing that help them better detect their pups’ calls. Wendy Saltzman, a behavioral neuroendocrinologist at the University of California, Riverside, investigates whether male P. californicus exhibit similar transformations. “We’re looking at various regions of the auditory system. . .and also at behavioral responses to pup cries,” she explained. The aim is to uncover whether fatherhood alters their hearing to enhance detection of and attention to pup vocalizations. Preliminary analyses of unpublished electrophysiological data from the auditory cortex hint that fathers may indeed process sounds—particularly pup calls—differently than nonbreeding males, she revealed.  

Although a mechanism discovered in a specific species, such as P. polionotus or P. californicus, cannot be directly extrapolated to other mammalian examples of biparental care, these findings can certainly offer clues about what to look for in other species, said Saltzman. “[These mice] present a much more natural model than, say, house mice, which are really used a lot for studying paternal care even though that’s not what they naturally do.” Thus, Saltzman added, “What we learn in the lab [from these Peromyscus monogamous and biparental species] is probably more relevant to what’s going on in nature.”

Flee or Freeze: Decoding the Brain’s Defensive Choices

Baier first started working with Peromyscus mice when he joined the Hoekstra lab as a graduate student in 2014. Around that time, a paper on defensive behavior in Mus musculus caught his attention. The study reported that lab mice attempted to escape when an expanding circle on a computer screen, simulating the looming approach of a bird of prey, was placed above them.¹⁴ This was surprising, given that neither these mice nor their ancestors have seen a predator for many generations, Baier noted.

The observations prompted Baier to replicate the experiments with various Peromyscus species. While burrowing, mating, and parenting behaviors in Peromyscus mice have been extensively studied, their strategies for responding to threats had largely been overlooked. 

“We built an arena, and I tested a bunch of different species,” he explained. He examined several strains derived from a diverse set of lab colonies in search of variations in their responses to these fake predators. He found significant differences between a P. maniculatus strain that originated from areas in Michigan with dense, thick vegetation and a P. polionotus strain from Florida’s open fields. P. maniculatus responded more quickly, often escaping at the sight of the threatening shape, while P. polionotus initially froze before eventually running away, reflecting perhaps adaptations to their distinct environmental contexts.

To identify the neural circuits behind these responses, Baier and his colleagues first looked at the activity of neurons in the superior colliculus, a midbrain structure that integrates stimuli from both eyes and ears. They found no differences between the two species. “That all pointed to further downstream in the brain,” said Baier.

Neurons in the superior colliculus connect to the dorsal periaqueductal gray (dPAG), a brain structure known to control defensive behavior and initiate escape.¹⁵ The team looked at neural activation in this region and it was there where they did see a difference. Electrophysiological recordings revealed that neurons in the dPAG of P. maniculatus were more active during the defensive behavior compared to P. polionotus. Additionally, the activity of these neurons correlated with escape speed, Baier and his colleagues reported in a preprint.¹⁶ 

“It’s quite interesting because you would think that that the easiest way to change these behaviors is just by changing the input, but that’s not what we found,” said Baier. “Maybe evolution has more ways of changing and acting in the brain than we previously thought,” he added.

For Tiago Branco, a neuroscientist at the Sainsbury Wellcome Centre who did not participate in the study, learning about where evolution fine-tuned this response in the brain was one of the study’s most interesting insights. 

Baier also emphasized the importance of studying behavior in animals with natural variation. Many labs study neurological diseases, such as those related to fear, using genetically identical lab mice, an approach that does not always reflect the natural human diversity underlying some of these diseases, he said. 

Carmen Sandi, a behavioral neuroscientist at the Swiss Federal Technology Institute of Lausanne who was not involved in this research but studies fear in different rodent models, said that findings like this are a good starting point for developing hypotheses. Yet, she cautioned that it is difficult to generalize principles from only two species. While this helps identify potential important brain modules for these behaviors, the adaptations described may be very specific to these strains.

But she is also enthusiastic about the neurobiology insights gained from these nontraditional models. “At the end of the day, we want to understand the nervous system, how it connects to function,” she said. Studies like this one could, in the future, contribute to a better understanding of the human condition. 

Branco added, “The work in neurobiology has been really dominated by [lab] mouse, because we have lots of genetic tools that we can use to understand how the brain actually works.” But now that there are more tools to study the brain of other rodent species, it’s extremely valuable to do so. “They give us a different perspective on how evolution has shaped circuits to generate behaviors.”