Disease Models

Autism in Mice: Developing Behavioral Standards for Mice Models

By January 6, 2016 August 27th, 2017 No Comments

Autism mouse models have been difficult to create for two primary reasons: 1) Genetic association studies show associations with genes involved in synaptic processing, but no autosomal dominant mutations are known to cause autism in mice and 2) The phenotype is not only varied (“on the spectrum”) but is subtle and difficult to test in rodents, whose cognitive processing abilities and social interactions are more limited than those of humans.

However, there are a variety of ways to test autism-like behaviors in rodents, and efforts have been made in the field to standardize such processes across the multitude of autism mouse models. Autism mouse models consist of knockout mutations of synaptic genes or knock-ins of human mutated genes that associate with the disease; often these models replicate disease phenotype incompletely. Because autism in mice is not a disease of intellect or memory, but rather of social and behavioral deficits, “autistic” mice are exposed to a battery of tests that differs from the set of memory and spatial relational tests commonly conducted to test learning across other disease paradigms. Efforts to standardize across these battery of tests and select among those most relevant to the understanding the disease have been carried out in recent years. Standard tests for autism in mice can be broken into three categories: social interaction, communication and repetition of behavior, all characteristics of the human disorder.

Autism in Mice Models: Testing

In assays of social interaction parameters such as: “nose-to-nose sniffing, nose-to-anogenital sniffing, following, pushing past each other with physical contact, crawling over and under each other with physical contact, chasing, mounting and wrestling” are measured in group social settings (“reciprocal social interactions”). A more targeted measure of social interaction is carried out in a three-chamber assay where an “autistic” mouse is given free range to interact with a novel object and a novel mouse, which is constrained to one chamber (“social approach” ). The three-chamber assay, as well as a Y-maze, can also be used to test an animal’s recognition of familiar versus novel mice (“social preference test”). The mouse’s degree of sociability is scored according to the amount of time the mouse spends with the novel mouse over the novel object, or novel mouse over familiar mouse. In both paradigms a normal mouse will spend more time in the chamber with a novel mouse.

Communication paradigm tests consist of exposing mice to pheromones, often from urine of other mice, and scoring their interaction compared to a non-biological scent, and from measures of ultrasonic vocalizations, that occur in normal mice when pups are removed from the nest, in resident females in a resident-intruder task, and in males in response to female pheromones.

Finally, assays of stereotyped behavior can be as simple as observing the mouse for repetitive circling, grooming and jumping and restricted exploration of an open field. Test of repetitive behavior may also include reversal learning tasks. This last category of tests utilizes familiar tests like the T-maze and Morris Water Maze, but in this case, the animal is test on its ability to re-learn flexibly, to find a food reward in the opposite arm of the T-maze from where it first learned to find the reward, for instance.

With an established set of tests for autism in mice, researchers have the means to compare new mouse models against existing ones in a comprehensive way and better ways to assess therapies. Recently, researchers have linked autism in mice with irregularities in certain areas within the chromosomes and begun to develop more complex and realistic models, in which environmental factors, combined with genetic risk, result in disease. Such models offer promise for testing behavioral and drug therapies that aim to ameliorate or reverse the effects of negative environmental stimuli on at-risk populations.


  1. Silverman JL, Yang M, Crawley JN (2010). Nature Reviews Neuroscience. PMID: 20559336 http://www.nature.com/nrn/journal/v11/n7/pdf/nrn2851.pdf


  1. Takahashi T, Okabe S, Broin PO, Nishi A, Ye K, Beckert MV, Izumi T, Machida A, Kang G, Abe S, Pena JL, Golden A (2015). Molecular Psychiatry. 15: 1359-4184 http://www.nature.com/mp/journal/vaop/ncurrent/pdf/mp2015190a.pdf

About Cailey Bromer

Cailey Bromer is a Neuroscience graduate student, science writer and aspiring data scientist. She regularly publishes on NeuWriteSD.org, as well as HippoReads.com, and has been contracted for various independent science writing projects. As part of her PhD work, she is teaching herself to program and develop code in various languages. Her work in Neuroscience revolves around understanding synaptic plasticity at the molecular level and how it contributes to larger computations in the brain