The complex mechanisms that underlie the onset of autism (autism spectrum disorders; ASD) are not fully understood, and researchers studying the neurodevelopmental disorder commonly use animal—especially mouse—models. A strain of mouse known as BTBR/J represents the most commonly used murine model of idiopathic autism. An international research collaboration headed by Kobe University’s Takumi Toru, PhD, and researcher Chia-wen Lin, PhD, has now made new discoveries that help to describe the onset of autism in mouse models.

In BTBR autism model mice (BTBR/J and BTBR/R variants), retrovirus activation makes copy number variations occur easily. In other words, they could be said to evolve at a faster speed than normal mice. The supporting evidence for this is that although BTBR/J and BTBR/R mice share a common ancestry, a mere 30 years of being raised in different environments has led not only to behavioral differences between the two species but also significant differences in brain structure (i.e., BTBR/R mice have a functioning corpus callosum, while BTBR/J mice do not). [Lin, CW., Ellegood, J., Tamada, K. et al. An old model with new insights: endogenous retroviruses drive the evolvement toward ASD susceptibility and hijack transcription machinery during development. Mol Psychiatry (2023).]

The combined results of detailed imaging and behavioral experiments using the BTBR/J mouse model, and a sister murine strain, BTBR/R, suggested that endogenous retrovirus (ERV) activation increases the susceptibility of a fetus to autism. The researchers also discovered that BTBR/R mice exhibit autistic-like behaviors without reduced learning ability, which they suggest makes this strain a more accurate model of autism than the widely used BTBR/J model.

It is hoped that further research will contribute toward a better classification of autism types, as well as point to new treatment strategies for neurodevelopmental disorders. Reporting on their work in Molecular Psychiatry (“An old model with new insights: endogenous retroviruses drive the evolvement toward ASD susceptibility and hijack transcription machinery during development”), the scientists said their study “unravels the idiopathic etiology of the BTBR strain by suggesting it as a superimposed model of multiple genetic mechanisms and virus infection.” They further suggested that targeting enhanced ERV activity or its infection process in BTBR mice will be the next step on the road toward developing possible therapeutic strategies for ASD of immune-dysregulated subtypes.

“ASD is a neurodevelopmental disorder with complex genetic architecture and heterogeneity, which have impeded the development of therapeutic strategies for this disease with a 2.3% prevalence rate in 2018,” the team noted. Reasons for this continuing increase in people diagnosed with autism include changes to diagnostic criteria, and older fathers becoming more common.

Autism is strongly related to genetic factors and can be caused by abnormalities in DNA structure, such as copy number variations (CNVs). “… accumulating evidence suggests a prominent role of CNV, large-scale structural variations in the chromosome, contributing either directly to ASD pathology or ASD susceptibility,” the investigators added.

Animal models, especially mice, are often used in research to illuminate the pathology of autism. Among these models, BTBR/J is a commonly used mouse model of the natural onset of autism. Studies have reported various abnormalities in BTBR/J mice, including impairment of the corpus callosum (which connects the left and right hemispheres of the brain) and excessive immune system signaling. However, it is not fully understood why this particular lineage displays autistic-like behavioral abnormalities.

The aim of the newly reported study was to shed light on the onset mechanism of autistic-like behavioral abnormalities. To do this the researchers carried out comparative analyses using BTBR/J and BTBR/R animals.

The investigators first conducted MRI scans on BTBR/J and BTBR/R mice to investigate structural differences in each region of the brain. The results revealed differences between BTBR/J and BTBR/R mice in 33 regions, including the amygdala. A particularly prominent difference was that while the brain corpus callosum in BTBR/J animals is impaired, in BTBR/R animals this part of the brain is normal. The corpus callosum is a region of the brain that connects the left and right hemispheres.

The research group next used array-based comparative genomic hybridization to compare CNVs in BTBR/R and normal, B6 mouse models. The results revealed that in comparison with B6 mice, the BTBR/R animals had significantly increased levels of endogenous retroviruses (ERVs). “ERVs are remnants of ancient retrovirus infection in the germline, the team explained. “Intriguingly, by analyzing the repeat sequences in the identified CNV, we found the potential involvement of ERV in speeding up CNV formation in both BTBR strains,” they stated.

Subsequent qRT-PCR analyses showed that these retroviruses were activated in BTBR/R mice, with comparative tests in the B6 animals indicating that this retroviral activation is specific to BTBR. Single-cell RNA analysis of BTBR/R mice revealed changes in the expression of various genes (including stress response genes) that are indicative of endogenous retrovirus activation.

Single-cell RNA analyses of embryonic BTBR mouse tissues provided further evidence of ERV activation in these animals, as expression changes were observed in a group of genes downstream of ERV. “… by single-cell RNA sequencing (scRNA seq), evidence of ERV activation during embryonic development was identified,” they further noted.

And while BTBR/J and BTBR/R mice have the same ancestry, experiments revealed differences in spatial learning ability and other behaviors between the two types of model mice. The researchers carried out a set of experiments to comprehensively investigate the differences between BTBR/J and BTBR/R at the behavioral level. “A behavioral test battery was performed to compare BTBR/R to BTBR/J using B6 as control,” they stated.

The results from these tests demonstrated that BTBR/R mice were less anxious than BTBR/J animals and showed qualitative changes in ultrasound vocalizations, which are measured as a way to assess communicative ability in mice. BTBR/R mice also exhibited more self-grooming behaviors and buried more marbles in the marble burying test. “BTBR/R spent ~30% more time on self-grooming than BTBR/J mice. In the marble-burying test, BTBR/R showed an even stronger repetitive behavioral phenotype with more than 75% of the marbles completely or half buried.” These two tests were designed to detect repetitive behavioral abnormalities in autistic individuals. From the results, it was clear that BTBR/R mice exhibited more repetitive behaviors (i.e., they were more symptomatic) than BTBR/J animals.

The 3-chamber social interaction test, which measures how closely a mouse will approach another mouse, also revealed more pronounced social deficits in the BTBR/R than in the BTBR/J mice. In addition, the results of spatial learning tests in the Barnes maze indicated that while BTBR/J mice exhibited reduced learning ability compared to B6 (normal mice), the BTBR/R mice exhibited similar abilities to those of the normal B6 animals. “… when hippocampus-dependent memory was analyzed by the Barnes maze, BTBR/R tended to learn the task faster than B6 and had similar latency to the goal at the end of the training, while BTBR/J demonstrated poor performance across the entire training period.”

Overall, the study revealed that retrovirus activation causes the copy number variants in BTBR mice to increase, leading to the differences in behavior and brain structure seen in BTBR/J and BTBR/R mice. “Intriguingly, disturbed epigenetic silencing mechanism leads to hyperactive ERV, a mobile genetic element of ancient retroviral infection, which increases de novo CNV formation in the two BTBR strains,” they stated. “This feature makes the BTBR strain a still evolving multiple-loci model toward higher ASD susceptibility … The ancient viral infection and reactivation affect host genome instability in the long term and have a continuing effect on embryonic development.”

BTBR/J mice are widely used by researchers as a mouse model of autism. The newly reported study results highlight the usefulness of the other lineage of BTBR/R mice because they exhibit autistic-like behavior without compromised spatial learning ability. “With the new advance in this old model, our study provides insights into how ASD susceptibility evolves in the genome and suggests BTBR/R as a precise model to investigate the core etiology of autism,” the team pointed out. The results also suggest that it may be possible to develop new treatments for autism that suppress ERV activation. However, it will be necessary to classify autism subtypes according to their onset mechanism, which is a vital first step toward opening up new avenues of treatment for autism.

a. Brain structure comparisons of the following mice: Left: BTBR/R and B6 (normal mouse), Center: Comparison of BTBR/J and B6, Right: BTBR/J and BTBR/R. b. Diffusion tensor imaging to compare differences in nerve fibers. Red indicates the brain regions that were either bigger or had increased numbers of nerve fibers in BTBR/J mice in comparison to either B6 (left and center images) or BTBR/R (right image). Conversely, blue indicates brain regions in BTBR/J mice that were comparatively smaller or had decreased numbers of nerve fibers. These scans revealed particularly significant differences between BTBR/J and BTBR/R mice’s corpus callosum. [Lin, CW., Ellegood, J., Tamada, K. et al. An old model with new insights: endogenous retroviruses drive the evolvement toward ASD susceptibility and hijack transcription machinery during development. Mol Psychiatry (2023).]