It has long been a topic of debate and discussion whether the lop-sided rates of autism diagnosis (males > females) reflects real gender variations in phenotype or simply diagnostic bias. Ozonoff et al. (2011) reported that males are three times more likely to receive an autism diagnosis as compared to females. However, Russell et al. (2011) that same year showed that females with similar levels of symptom severity were nevertheless less likely to receive an autism diagnosis than their male counterparts. This and similar studies have led many to wonder whether the diagnostic criteria, having been based primarily on male phenotypes, fall short in identifying affected females. Even Hans Asperger himself originally believed that “Autistic Psychopathy,” as he called Asperger’s Syndrome, occurred solely in males. (He later revised his opinions, however, having run across some cases of females who presented with the condition.)

Hans Asperger working with a young boy diagnosed with “Autistic Psychopathy,” now known colloquially as Asperger’s Syndrome or, professionally, as a high-functioning form of Autism Spectrum Disorder.

But a recent study published this last month by Jacquemont et al. (2014) supports the notion that, while diagnostic bias may well exist, some of the gender bias we’re seeing may be due to a “female protective effect” on overall genetic burden of deleterious mutations associated with autism. In short, those females in the recent study with an autism diagnosis were more likely than their male counterparts to house very large (> 400 kb) copy number variations (CNV) in autosomal chromosomes. (The sex chromosomes have an obvious gender bias and were removed from initial analyses to control for this fact.) A similar, though slightly less extreme, trend was seen in individuals with various developmental and intellectual delays. What this may mean is that males with lesser genetic burden may exhibit more severe phenotypes and that for a female to exhibit comparable severity, her mutations must be that much more severe to cross a similar threshold.

The team studied thousands of gene array data samples from individuals with developmental delay (DD), intellectual disability (ID), or autism. They studied CNVs (large and small) and single nucleotide variants (SNV). What the researchers found was that small or rare CNVs were equally distributed across both genders; however, CNVs larger than 400 kilobase pairs in length showed a distinct gender bias, with females exhibiting a heavier burden. The same was true when de novo or new CNVs were analyzed separately. When looking at all DD, ID, and autism patients together, females showed a remarkable 2-fold increase over males in the number of large CNVs. Even more remarkable, when the researchers looked at the data of autism patients alone, females exhibited a 3-fold increased burden of large CNVs. When the team studied rare truncating SNVs (mutations which yield a shorter, functionally-altered gene product) a similar though lesser trend was noted across all patients.

Although there was a significant inverse relationship in overall cognitive performance, e.g., IQ test, and the severity of genetic burden in females in this study, once cognitive variations were statistically corrected for the relationship still remained between gender and genetic burden. So the relationship between symptom severity and genetics appear to be more complicated than originally thought and perhaps may lie in variations within the different genes that are targeted in the female group.

One of the most interesting findings of this study, aside from the primary results, is the fact that it seems maternal inheritance for deleterious mutations is significantly higher for large CNVs. This was true for both the larger cohort including all DD, ID, and autism patients as well as the autism simplex patients alone. Within the autism simplex cohort, small SNVs and, most interestingly, large CNVs contained within genes not associated with any neurodevelopmental conditions showed balanced inheritance from mother and father. In addition, across all three groups, SNVs showed a maternal inheritance bias in neurodevelopmental and brain-associated genes. What this suggests is that if there is a female protective effect in DD, ID, and autism that it is brain-specific. Genes which were not associated with these conditions showed no apparent gender bias in terms of inheritance.

Why in the world would being female protect against the effects of mutations? Well, one thing which I find curious is that there are more of these mutations coming from the mother’s side. Mind you, this doesn’t necessarily mean that more maternal genes are being passed on versus paternal genes. What this smacks of, to me, is parent-of-origin effects or genomic imprinting. When a given mutation comes from the mother, a brain-specific mutation, it may be preferentially expressed over that of the inherited copy of the father’s. So when a mutation in the same gene is inherited from the father’s side, it may be less likely to cause a condition like autism. Recent work on genomic imprinting studies in mouse brain may bear this out. Gregg et al. (2010) reported that they observed an overwhelmingly maternal contribution to gene expression within the developing brain, meanwhile more genes from the father were expressed in adulthood. Since neurodevelopmental conditions arise due to problems in “development”, it would make sense that maternally-inherited genes would play a larger role in the phenotypes we observe. (This, however, does not mean that no paternally-inherited mutations play roles in autism or other neurodevelopmental conditions; there is simply a maternal bias.)

It may also mean that females who inherit mutations from their fathers may serve as the perfect silent carriers, since said females would preferentially express more of their maternally-derived brain-specific genes, leaving them less seriously affected by said mutations. They may, however, then pass down those mutations to the next generation, since their genes would indeed be preferentially expressed in the brains of their developing offspring. This may help explain the preference for maternally-inherited genes in neurodevelopmentally-affected individuals.

However, at least at first glance, this doesn’t necessarily explain why females who are affected have a higher genetic burden of deleterious mutations. Presumably the rates of inheritance of mutations from the maternal side are generally similar, except females require more severe mutations in order to cross symptomologically into the diagnostic ranges. Therefore, parental imprinting aside, there may well be something about the development of the female brain which reduces the deleterious effects of these types of mutations, up to a certain point. Hormones, after all, play varied and integral roles in aspects even of the earliest of brain development.

It will be interesting to see how this research continues to develop, especially regarding sex-specific mechanisms of protective action. In addition, it’s intriguing to consider that many neurodevelopmental conditions may be maternally inherited because of the roles that maternal imprinting plays on early brain development.