Autism and the Male Brain

A new study reveals an early divergence of male and female brains.

Posted Jan 30, 2021 | Reviewed by Gary Drevitch

The question of whether there are male brains and female brains is a subject of considerable debate in both neuroscience and gender studies. While some researchers use neuroimaging and other tools to identify differences in brain structure and function between men and women, others believe that the search for such differences is itself nothing more than an attempt to reinforce perceived male and female roles and status. Serious issues lie beneath this discussion, none more so than the differential prevalence of some medical conditions between men and women. More women than men are diagnosed with multiple sclerosis or anxiety; more men live with ADHD. If male and female brains aren’t different, how can this sexual dimorphism be explained?

Autism Spectrum Disorder (ASD) is a particular case in point. It has long been thought more prevalent in boys. When Leo Kanner first described the condition at Johns Hopkins Hospital in the 1940s, eight of his first 11 patients were boys. Similarly, the first four patients described by Hans Asperger in Vienna were boys. This prevalence in males has persisted and been confirmed by more recent studies. The consensus among psychiatrists now is of a roughly 4 to 1 ratio. Less clear is whether this is a real difference, or rather a result of ascertainment bias because the condition is less well recognised in girls. Perhaps psychiatrists don’t anticipate autism in girls, so they don’t look for it. Certainly, the symptoms appear different in girls compared with boys. Where boys show less social behaviour and more aggression, girls show more depression, anxiety, and emotive behaviour. Of course, this difference itself speaks to a brain difference between the sexes, but if diagnostic criteria need to be different for different groups of patients, then getting the diagnosis right becomes even more tricky than is already the case for this complex condition.

The idea that autistic girls are being overlooked is given added credence by the absence of convincing biological explanations for the discrepancy. Beyond the idea that boys simply have more testosterone, why should boys be more susceptible to whatever brain changes underlie the emergence of autism? A recent study [1] has addressed this issue directly, and an interesting answer has emerged.

The most powerful risk factors for ASD are genetic, but environmental factors are also important. One factor that has consistently caught the attention of autism researchers is ‘Maternal Immune Activation’. If a mother gets a serious viral illness during pregnancy, the risk that her child will subsequently receive a diagnosis of autism is raised significantly. The degree of increase depends on the virulence of the virus, but for a common virus such as influenza it is probably around three-fold. So as I’ve discussed in an earlier post, if Mum gets a severe bout of the flu, the child’s risk of an autism diagnosis increases from around 1-2% to perhaps 3-6%. (I won’t diverge at this point to consider the potential impact of our current scourge, the COVID-19 virus, having dealt with that in an earlier post.)

Interestingly, the increased risk is not the result of the action of the virus itself on the developing fetus. Rather, it is the inflammatory response the virus provokes in the mother that does the damage. We’ve all had enough colds and flu to recognise this inflammatory reaction: increased body temperature, headache, malaise, tiredness. These are caused by the action on the body of pro-inflammatory factors (called cytokines) released by the immune system into the bloodstream in response to the infection. Critically, during pregnancy some of these factors find their way across the placenta into the fetal bloodstream, from where they can interfere with brain development, pushing it towards an autistic outcome.

This phenomenon has attracted a lot of attention from neuroscientists for an obvious reason: If they could understand the consequence of this cytokine invasion, they might gain a better understanding of why autism arises. They have been aided by the fact that this effect can be reproduced in mice. 

If a pregnant mouse is injected with a substance called poly (I:C), then she mounts the same pro-inflammatory response that she would to a real viral infection. This is because poly (I:C) looks to the immune system just like a real virus, without actually being an infective agent. Remarkably, the mouse pups born following a poly (I:C) inoculation will show behavioural abnormalities that correspond in many particulars to the behavioural features of autism. The pups show altered social behaviour; they have repetitive behavioural patterns; they develop strange vocalisations. It would certainly be going too far to claim that these were autistic mice. Nonetheless, the correspondence with autism is remarkable.

So far, nothing new: This has all been known for years. But this new study shows that male mice are more susceptible to this effect than female mice.  This demonstration involves some pretty high-powered molecular biology. The researchers focus specifically on the action of one cytokine called Il17a. Previous research had indicated that this was a pivotal pro-inflammatory factor in inducing the behavioural changes observed in the mouse pups. Using a technique called single-cell transcriptomics, the team analysed individual cells in the brains of the mouse fetuses following exposure to this cytokine in utero. They discovered changes in gene expression; that is, these brain cells had turned on some genes that they wouldn’t normally express, and had turned other genes off. These gene expression changes led to increased activity in the individual cells deep in the developing cerebral cortex of these mouse fetuses as a consequence of this cytokine impact. 

But the remarkable observation was that this change only happened in the brains of male fetuses; female fetuses showed no such change. Female mice didn’t show the behavioural change associated with the Il17a exposure; they didn’t show the gene expression change; nor did they show the increased brain cell activity observed in the male mice. The inescapable conclusion is that at the point of Il17a exposure (a point roughly equivalent to 1st/2nd trimester in humans) not only were male and female mouse brains demonstrably different, but they were different in a way that impacted their susceptibility to subsequent behavioural outcomes.

What is the difference between the male and female brains? The study shows that the difference lies in what molecular biologists call the integrated stress response. All cells in the body have mechanisms to respond to stress. Remember, all cells in the world of biology are ultimately derived from free-living, single-celled organisms. These one-cell creatures had to evolve protective mechanisms; they had no immune system to defend themselves. Even now in complex multi-cellular organisms like ourselves, individual cells have retained these cell autonomous functions and react individually to threatening stimuli like temperature, oxygen deficits, of lack of nutrients. It transpires that male brain cells (in the mouse) activate this stress response to IL17a exposure while the female cells do not. The consequence for male cells is dysregulated gene expression and protein synthesis, and downstream, an increased susceptibility to the behaviours we call autism.

Why have male brains evolved this way? You might imagine that since evolution has managed to build a brain without this apparently detrimental proclivity (i.e. a female brain), it would utilise the same strategy when building male brains as well. If females have evolved a resilience mechanism, why haven’t males got it too? Is there some advantage for the male in retaining this seemingly disadvantageous feature? Another obvious question; Is this a unique mouse feature or is it shared by men? I think we can be pretty certain that experiments designed to answer that question are currently underway in a laboratory near you.

The outcome of this study that will probably most interest clinicians, carers, and persons with autism themselves is the observation that this effect of maternal immune activation is reversible. Drugs that block the stress response, administered after the initial IL17a administration, prevented the behavioural change emerging in the male mice. This raises the possibility that autism arising in boys via this mechanism could be blocked in the same way. Such a therapeutic approach, however, would be fraught with difficulties. We’ve noted already that roughly 5% of children whose mothers suffer a severe viral illness go on to receive a diagnosis of autism. This obviously means that 95% of such mothers have neurotypical children. To treat all mothers who suffer viral illness in pregnancy indiscriminately would surely be unethical, even if it proved safe, which is by no means certain. So unless a biomarker could be discovered that identified the 5% of pregnancies at risk, this is unlikely to become a practical proposition. Even so, the identification of such a biomarker may well follow on the heels of this study.

Good science always raises more questions than it answers, and this work is no different. Maternal immune activation is a relatively minor risk factor for ASD. Genetic factors are much more important. Could they also work via the same mechanism? There are notable overlaps: Some of the genes associated with autism impact protein synthesis and cellular trafficking, superficially similar to the impact of the integrated stress response identified here. Also, how does the difference between male and female cells arise? Human male and female fetuses begin to differ at about 5 weeks of gestation when male and female gonads begin to take divergent paths. Presumably, the ‘resilience to stress’ mechanism emerges in females downstream of this divergence, but how? Lots still to learn.  

It's interesting to note, finally, that studies of biological sex need not necessarily have a pro-male bias: This study clearly identifies a male vulnerability not shared by females.


1. Kalish BT, Kim E, Finander B, Duffy EE, Kim H, Gilman CK, et al. Maternal immune activation in mice disrupts proteostasis in the fetal brain. (2020) Nat Neurosci.