Mighty Mouse Models of Memory
Mice have become an important model of humans in medical and scientific research, because we can control both their nature and their nurture. But one aspect of many mouse studies that makes them nice and tidy also prevents the findings from translating well to people: Researchers tend to use ‘inbred’ mice, which are genetically identical to each other. That helps silence the noise caused by genes unrelated to those of interest, making it easier to isolate genetic or environmental factors that contribute to aging or other processes.
Humans, meanwhile, have all kinds of genetic makeups, which can influence how they respond to aging, disease, diets or medicine. For example, not everyone who has APOE4, a genetic risk factor for Alzheimer’s disease, develops the condition. Nor does everyone respond the same way to a change in diet, including caloric restriction, or to therapeutics for other brain diseases such as Parkinson’s and multiple sclerosis.
Studying only a single mouse strain is “like studying only a single human,” says Catherine Kaczorowski, a neuroscientist at the Jackson Laboratory (JAX) in Maine. “This is a major limitation if you want to generalize the findings to treat other mouse strains, and eventually humans.”
To try to understand how different genetic backgrounds can influence aging and responses to interventions, scientists have developed new, more genetically diverse populations of mice. “It’s much more complicated and expensive to do a project with genetic diversity,” says Kristen O’Connell, a biologist at JAX who works with Kaczorowski. “That said, it is becoming much more widely appreciated that it is important to include genetic diversity. So I think the whole field is undergoing transition.”
Kaczorowski’s lab has been a leader in this transition, focusing much of their work on genetically diverse strains and even developing new models for a variety of human age-related diseases. They have found that, like humans, mice with different genetic backgrounds age very differently, with some showing faster memory decline and others functioning well into old age. They are now using those mice to identify genes that influence cognitive aging and to uncover how those genes influence cells, circuits and, ultimately, memories.
“She’s been one of the very few top-flight labs over the last five years that has devoted much of her own research to documenting the importance of using genetically heterogeneous animals,” says Richard Miller, a pathologist at the University of Michigan who also studies aging using diverse mice.
Dudley Lamming, an endocrinologist at the University of Wisconsin-Madison, calls their work on genetically diverse mice and cognitive aging “pioneering” and says it has “opened the door to understanding the underlying causes of [cognitive aging] in the human population.”
Strength in Diversity
Cognitive aging in people clearly has a genetic component, but the degree and mechanisms are not clearly understood. Many studies suggest that intelligence is roughly 50 percent heritable through genes, depending on the age you’re tested. Fewer studies have been done on the rate at which performance declines as we age. Twin and family studies estimate that the rate of longitudinal decline is roughly 50 to 80 percent dependent on genetics, Kaczorowski says. (The genetic heritability of Alzheimer’s disease is up to about 80 percent.)
Some studies have looked at particular genes that might be responsible for differences in cognition late in life. The APOE gene codes for a protein that carries fats to neurons, and its alleles are associated not only with risk of Alzheimer’s disease but also with memory and executive function in healthy subjects. Genes involved in inflammation and oxidative stress also crop up in studies of genetic variance and cognitive aging. But identifying specific genetic variants, especially those tied to resilience to cognitive decline, has been challenging. Studies in mice, in which both the genomes and the environment can be controlled, could point to new targets.
In 2012, a team at JAX created the JAX Diversity Outbred (DO) mouse population, by crossbreeding eight inbred strains carefully selected to cover a wide variety of traits related to diabetes, cancer and other medical conditions. Kaczorowski’s team has since used DO mice to search for genetic variants tied to cognitive aging. In one study, researchers assessed working memory in DO mice aged 18 months, roughly equivalent to 60 human years, then mapped their genomes. The results, published in Cell Reports in 2020, showed that a particular variant of one gene, DLGAP2, strongly predicted memory performance in old mice. “What’s really cool about this gene is that we know that it has an essential role in neurons,” says Andrew Ouellette, a graduate student in Kaczorowski’s lab, who led the study with Sarah Neuner and Logan Dumitrescu. “It has been implicated in autism and schizophrenia, but not yet in Alzheimer’s disease or normal cognitive decline.”
DLGAP2 codes for a protein that affects the shape and structure of dendritic spines, which mediate communication between neurons. Previous research shows that the spines’ shape is linked to memory. Ouellette and collaborators found that mice that had performed well on a memory task had a lot of thin spines and few stubby spines, further implicating the gene in cognitive aging. They next looked at humans, comparing post-mortem brain tissue with cognitive performance at the last assessment before death. People with lower expression of the DLGAP2 protein in an executive area of their brain called the dorsolateral prefrontal cortex performed more poorly at this assessment. This was true among people diagnosed with Alzheimer’s disease, those with minimal cognitive impairment, and also those with normal cognition.
Ouellette says DO mice have been essential to his work. Just like humans — identical twins aside — every DO mouse is genetically unique. “We can encapsulate the genetic diversity we see in the human population, but study that in a controlled lab environment.”
Ouellette used DO mice again in a study just published in the October issue of Neurobiology of Aging examining the potential of dietary restrictions to preserve late-life cognitive performance. Previous research has found that caloric restriction and intermittent fasting improved both life span and memory in old age in mice. But Ouellette and collaborators revealed a more complex picture — genetic background shapes how restricted diets affect cognitive aging. When averaged across the mice, these interventions didn’t improve memory performance in 22-month-old mice. The diets may have improved memory in some mice, specifically those with certain variants of the SLC16A7 gene, but they worsened memory in others.
A Whole New Breed
Ouellette says he uses DO mice mostly for exploratory purposes, to find genetic markers for cognitive outcomes. “Which is hugely important right now, because as a field we understand so little about how the brain works,” he says. “We just need a fundamental understanding of these genetic mechanisms.”
But for some purposes, DO mice are actually too diverse — as a family, they carry roughly 50 million genetic variants, roughly on par with human genetic diversity. If you want to nail down the correlations between genetic patterns and health outcomes, you need to study a lot of DO mice to gain statistical power, because they’re all so different and only a few might have a given genetic pattern.
Kaczorowski and collaborators concocted a kind of Goldilocks solution by crossing B6 mice, a common strain of inbred mice, with a family of mice called BXD, consisting of nearly 200 genetically distinct strains. “[BXD mice] have been around for a long time, so there’s a wealth of legacy data on them,” O’Connell says. Not only do researchers understand their genetics well, but many popular genetic tools, such as CRISPR, were developed using these strains. So scientists know where they can, say, insert reporter genes — genes that might cause a cell to fluoresce when a gene of interest is active — without worrying about interfering with other cellular mechanisms. In comparison to DO mice, the new family, known as B6-BXD, has about 6 million variants. “That’s way more diversity than is typically included in mouse studies but 10 times less than in humans,” Kaczorowski says.
Because this panel of mice has an array of reproducible genotypes, the researchers can use them to explore how different background genetics influences the rate of normal age-related cognitive decline. In a paper published in Frontiers in Cell and Developmental Biology in 2020, they found that different strains performed differently on memory tests over time. By midlife, some strains were susceptible to cognitive decline, but some were not. The findings confirmed that cognitive decline is highly heritable and likely driven by a combination of many different genes. Researchers are now assessing memory at older ages, which they hope will reveal strains that are reliably resilient to cognitive decline.
Once researchers have identified genes tied to cognitive aging, they can use the B6-BXD panel to explore how those variants of interest actually affect the brain. “We can link genetic architecture to cellular and physiological traits,” Kaczorowski says. “We will have the power to figure out if something is a driver of cognitive aging versus something happening in parallel or a consequence.”
As part of a new project funded by the Simons Collaboration on Plasticity and the Aging Brain, Kaczorowski, O’Connell and Vilas Menon, a neuroscientist at Columbia University, will examine the functional and structural changes in neurons that express high levels of genes tied to cognitive resilience. Researchers will sequence RNA in single cells and record the cells’ electrical activity to get a sense of how genetic changes influence neural activity. “What the Simons grant allows us to do for the first time is look at directionality,” Kaczorowski says.
Menon will compare mouse findings to post-mortem tissue from human brains to determine which changes in mice might be translationally relevant to humans. “If we can align the compositions between the mouse cells and the human cells,” Menon says, “we can probably make some inferences about what the electrical activity of the human cell would have been like.”
From Cage to Clinic
The researchers have already learned a lot from their work on cognitive decline. O’Connell says she’s been surprised by how much even a little genetic variation can affect behavioral variation. “I think one of the other things that’s been really interesting is the extent to which just normal cognitive aging can explain decline associated with Alzheimer’s disease and other neurodegenerative diseases,” she says. “Which maybe shouldn’t be that surprising, because age is the biggest risk factor for those diseases. But rather than targeting some of the neuropathologies, if we actually target the biological processes of cognitive aging, we might be able to enhance cognition, even in diseases like Alzheimer’s disease.”
Along similar lines, Kaczorowski says researchers typically study cognitive decline and neuropathology in tandem, comparing individuals with both to those with neither. She’s interested in the other two quadrants — individuals with one but not the other. By separating memory loss from brain plaque, she’s found that the majority of cognitive decline results not from neuropathologies like beta-amyloid buildup, but from other genetic and environmental factors. There are some people, she says, who, “if you looked at their memory score, and you tried to predict what was in their brain, you would imagine that their brain looks like a mess, right?” she says, and yet their brains look fine. “So there’s this dichotomy between brain pathology and how someone really ages cognitively.” Identifying protective genetic factors that help keep memory intact despite brain pathology offers a promising route to preventing cognitive decline. (For more, see “Super Agers and Centenarians: The Search for Protective Factors.”)
Genetically diverse mice are helping with this search as well. Kaczorowski and collaborators have used a similar strategy to their B6-BXD mice to study genetic factors that influence the severity of Alzheimer’s disease, developing a BXD panel that carries a genetic variant tied to beta-amyloid plaques. Among mice that inherited the Alzheimer’s-related gene from their mother, some showed much less memory impairment than others did. That suggests that some of the genes they inherited from their fathers may have played a protective role. The existence of such protective genes would not have been as apparent without a diverse population of mice. The researchers are now trying to identify those genes.
Kaczorowski is already thinking about how to translate her work into future therapies — she has patented one gene that’s shown therapeutic potential and screened more than 100,000 molecules that target proteins of interest, narrowing the candidates down to one she’ll test for efficacy in mice. Testing candidate therapies in diverse mice will give researchers a sense of whether a particular treatment is broadly effective or works in only some populations, Menon says. “Thanks to people like Catherine and Kristen,” he says, the use of genetically diverse mice is “now something that the field is recognizing as a necessary tool.”