Not All Mutations in Autism Are Just about Loss- or Gain-of-Function
A few weeks ago I wrote about cell stress and its relationship with epilepsy. I reviewed how the accumulation of misfolded proteins within the endoplasmic reticulum (ER) can lead to an Unfolded Protein Response (UPR), which, if chronic, can lead to cell damage and even cell death.
This week, I’m talking about autism.
When we talk about “neurodegeneration”, probably the first thing that comes to mind is something like Alzheimer’s or perhaps even Huntington’s Disease. In these conditions, it’s obvious that there’s a progressive loss of cognitive function and we know that this is accompanied by neuron loss as well. Autism, on the other hand, while it may experience some setbacks in the form of regression, typically follows the opposite progression: it improves over time. So at first glance, one would think that end-of-life neurodegenerative diseases have little in common with a neurodevelopmental condition like autism. However, there may in fact be more there than meets the proverbial eye.
For the last decade, the scientific community, and particularly the geneticists, have been enraptured by potential synaptic involvement in the autism phenotype. There have been a number of candidate risk genes whose gene products function at the synapse (although they may also function elsewhere in the cell), and because of this, enthusiastic scientists have had a tendency to cherry-pick the genetic data to focus on the synapse in autism. (At one point I gave them a greater benefit of the doubt, however my recent work in genetics indicates that there are a comparatively small number of genes in autism that are involved strictly at the synapse and that risk genes tend to affect a broader range of neural development than typically acknowledged. So, um yeah, definitely guilty of some serious cherry-picking.)
In the study of neurodegeneration, it is well-acknowledged that not only may a particular genetic loss- or gain-of-function be responsible for a given condition, but the mutation may lead to a change in the shape of the gene product when one is produced. (This is especially more likely when small mutations such as point mutations occur, rather than a large deletion encompassing a gene.) In these instances, a change in shape can lead to stalled protein processing within the lumen of the ER, causing an accumulation of that gene product. Not only does this mean that the cell is likely haploinsufficient for that product (it only produces half as much as a normal cell), which may or may not impair cell function, but the build-up of protein within the ER can lead to cell stress, triggering the UPR, and can lead to altered physiology and even cell death if chronic or severe enough.
Interestingly, there are a number of studies in autism that suggest this may also be the case, although strangely they have received little press within the scientific community. Back in 2010, Fujita et al. reported on several cases of point mutations within the Cell Adhesion Molecule-1 (CADM1) gene. They found that function may have been only modestly reduced (~20%), suggesting a minor loss-of-function. However, they also noted that the mutated Cadm1 protein exhibited abnormal intracellular accumulation. When they investigated its specific intracellular location, they found that it was abnormally accumulating within the ER. It was also colocalized with a molecule known as beclin, which tends to localize within the ER under ER stress. They concluded that the mutated protein led to its abnormal accumulation within the ER and subsequent activation of an ER stress response, i.e., UPR.
What I find particularly interesting is that, even though the team’s experiments suggested there was only a minor loss-of-function in terms of Cadm1 activity, neurons transfected with the mutant protein exhibited short dendrites or failed to develop dendrites at all. So either a mild insufficiency of Cadm1 is unusually detrimental to dendrite development OR the phenotype is complicated and exaggerated by the ER stress response, the latter of which makes more sense given the UPR’s involvement in regulation of various membrane receptors, such as glutamate and GABA receptors, as well as protein translation and cell growth in general.
Interestingly, several studies have reported similar findings on Neuroligin-3 and Neuroligin-4, indicating that these point mutations likewise lead to the accumulation of their protein products in the ER. Though they’re currently few in number, these studies highlight the fact that ER stress may be an integral part of some individuals’ phenotypes, and may not only help us understand the cause but may also offer some means for amelioration of symptoms.
It also may lead one to wonder how much the loss- or gain-of-function of the actual protein product itself is involved in the individual’s condition or whether the ER stress may itself be a bigger driver of autism risk.
Some evidence suggests that could be the case (although when is biology ever so simple?). For instance, it has been suggested that the UPR can affect and may play important roles in neuronal differentiation, a stage of development I’ve considered far more integral to autism risk [1, 2]. In addition, Momoi et al. (2010) have already proposed that
“[ER stress] arising from these mutations causes a trafficking disorder of synaptic receptors, such as [GABA B-receptors], and leads to their impaired synaptic function and signal transduction. . . . [We] propose a hypothesis that ASD pathogenesis is linked not only to loss-of-function but also to gain-of-function, with an ER stress response to unfolded proteins under the influence of epigenetic factors” (p. 13).
As an addendum to the above statement, I would propose that ER stress affects not only synaptogenesis but likely all stages of neuronal development, as we have been finding.
Again, it is shocking that this work by Fujita and colleagues has not received more attention by mainstream autism science. And yet this area probably holds the greatest promise for treatment intervention. Already, there are a number of medications on the market used to treat neurodegenerative diseases, some as simple as over-the-counter aspirin. Why then has this work been ignored, especially considering the platforms from which the evidence has been shouted, like Nature and The Journal of Neuroscience? Why has this grabbed so few scientists’ attentions? The implications are astounding for genetics research.
Perhaps scientists just don’t get how astounding.