Autism isn't just about genes & proteins; it's also about how those proteins are folded.

By Diana Crow

Biology is never simple. Geneticists have identified dozens of genetic mutations that are more common in autistic people than in the general population, but researchers like University of California- San Diego psychiatric systems biologist Lilia Iakoucheva are quickly discovering that genes are only part of the story.

When geneticists collect samples to analyze, they rarely take samples from neural tissue. When researchers describe a negative effect of a particular mutation, they're often describing an effect they observed in a blood cell or a skin cell and assuming that the same effect happens in neural cells as well.

Iakocheva wondered if that assumption was safe, so she teamed up with a group of scientists at the Dana Farber Cancer who study protein interaction networks to find out.

Image Courtesy of Argonne's Midwest Center for Structural Genomics and Wikimedia Commons

---

Comments? Leave them below!

If DNA is the blue print, then protein interaction networks are the maintenance reports that describe how well the systems in the building are working. DNA tells the cell how to build a wide variety of proteins, the small organic molecules that make up the bulk of the cell, but there are a lot of factors that can affect which proteins are made, how many copies the cell makes of a protein, and how proteins are shaped in their final form.

Protein shape is crucial for proper function, since most cellular processes depend on proteins fitting together like cogs and gears in an old-school pocketwatch. Even slight variations in protein shape can have a huge impact on how the cells behave.

To understand what is happening at the level of the organism, scientists need to understand not just the genetic code but also the path leading to the end product.

Building this understanding is crucial, because the goal of genetic research into autism and other neurological conditions is to develop medications that specifically target the proteins that behave differently in atypical neurons.

Up until this decade, developing psychiatric medications has been a process of trial and error. Psychiatrists prescribe stimulants and anti-depressants, because they appear to work, based on patients' outward behavior and self-reports. But we know very little about the biochemistry behind why people respond to psychiatric medications the way they do.

This trial-and-error process makes many researchers, including Iakocheva, uncomfortable.

The problem is that the path of protein production varies a lot in different types of cells. For example, neurons have to be able to send and receive electrical signals much more efficiently than other cell types, so they produce the proteins that deal with electrical signals in higher quantities than other cell types. Conversely, muscle cells have to produce larger amounts of the stringy proteins that allow them to expand and contract.

Because the concentrations of proteins vary depending on the cell type, an interaction that occurs frequently in blood or muscle cells, may be extremely rare in a neuron. Alternatively, an interaction that causes problems in neural cells may not occur in other tissue types simply because the two proteins causing the problem don't cohabitate in non-neurons.

Keeping track of all these proteins is a no simple task – especially when taking into consideration all the alternate versions of the same protein – but scientists from all over the world are collaborating to build databases of proteins that interact with each other.

That's where Iakoucheva's work comes in. She studies the protein variants that occur in neurons and how slight variations in protein-assembly can alter how the proteins work in neurons from autistic people's brains.

However, even the most definitive protein-tracking databases only look at whichever protein shapes the scientists found in the cells that were easiest to study. The lack of tissue-specific information about protein interactions may be causing scientists who work on tissue-specific conditions to miss important details. Iakoucheva and others are concerned that without looking at the tissue-specific isoforms, researchers won't be able to develop medications that address what's happening in autistic people's neurons.

So she decided to build two maps of protein interactions, one based on the protein isoforms that are already in the databases, and another based on isoforms that have been observed in neurons. Because it would be almost impossible for one lab to build these maps on its own, she teamed up with proteomics researchers at the Dana Farber, who have built similar protein maps for cells affected by a wide variety of diseases.

Half of the protein interactions they observed only occurred when previously unstudied versions of the protein were in the mix. They also found that some of the protein interactions the databases predicted did not occur between the protein isoforms that actually live in neurons. The inevitable conclusion was: We need more research on protein variation.

“The genes that we are discovering, of course, are very important. But ultimately, we have to go a step further to develop drugs,” said Iakoucheva. “You have to know what the proteins are. Focusing on those that are more related to the brain can shorten the pathway to drug development...hopefully.”

References:

Corominas R*, Yang X*, Lin GN*, Kang S*, Shen Y, Ghamsari L, Broly M, Rodriguez M, Tam S, Trigg SA, Fan C, Yi S, Tasan M, Lemmens I, Kuang X, Zhao N, Malhotra D, Michaelson JJ, Vacic V, Calderwood MA, Roth FP, Tavernier J, Horvath S, Salehi-Ashtiani K, Korkin D, Sebat J, Hill DE, Hao T, Vidal M, Iakoucheva LM. Protein interaction network of alternativelyspliced isoforms from brain links genetic risk factors for autism. Nat. Commun. 2014 April 11;5:3650 doi: 10.1038/ncomms4650

Kitchen RR, Rozowksy JS, Gerstein MB, Nairn, AC. Decoding Neuroproteomics: integrating the genome, translatome, and functional anatomy. Nat. Neuroscience. 2014 October 28; 17:1491-1499. doi:10.1038/nn.3829

Yeo G, Holste D, Kreiman G, Burge CB. Variation in alternative splicing across human tissues. Genome Biology. 2004 September 13; 5:R74 doi:10.1186/gb-2004-5-10-r74

About the Author: Diana Crow is a freelance science journalist, living and working in the city of Boston, MA. She has written for print and web publications including Scientific American, Minority Postdoc, and Method Quarterly, among others. She is also a project manager and science reporter for The National Science Communication Institute‘snSCI Profile Series and co-editor-in-chief of the Scientista Foundation blogs.

*Note: The author of this post is involved in the neurodiversity community, a loose network of people who subscribe to the idea that neurodevelopmental variants like autism, ADHD, dyslexia, etc. are an integral part of human diversity. Like many others in the community, this author frequently argues that while many people with atypical neurological wiring may need assistive devices, alternative communication strategies, and/or medications in order to live their lives to the fullest, they still have the right to decide what treatments, interventions, & accommodations they receive.

The neurodiversity community is also concerned that many of the pathology and deficit-based terms neurotypical people use to describe non-neurotypical people often have negative impacts on non-neurotypical patients' self image and confidence. Needless to say, there's a lot of variation in beliefs within the neurodiversity community. This author maintains the position that biomedical research and knowledge can be incredibly useful and empowering, as long as it's applied to patients in a humane and consensual way.