Emerging technologies have the potential to transform our understanding of neuropsychiatric and neurodevelopmental disorders. It is likely that rare variation and new mutations in many genes play significant roles in the risk of developing these disorders. Technical advances and next-generation sequencing have allowed us for the first time to sequence the ~1% of the human genome that contains protein-coding genes, termed the 'exome', in many individuals. Using exome sequencing, we can now identify most of protein-altering variants present in an individual. However, predicting which of these ~20,000 variants are contributing to disease remains challenging, even when the likely risk genes are known. In fact, humans have a significant background of rare variation and computer models still do a poor job of predicting which variants will be truly disruptive. Importantly, these variants are so rare that they may be restricted to a single family, meaning they will have never been seen clinically. In addition, many genes have now been implicated in a wide variety of diseases and it is unclear how different mutations in the same gene lead to such diverse clinical outcomes. Understanding the functional effect of a patient's mutation is a key part of developing personalized medicine.
We propose to develop new technologies that allow testing for a gene of interest--here PTEN--the functional effect of all possible mutations. Recent evidence from exome studies suggests that rare and spontaneous new mutations in the PTEN gene play a major role in cancers, overgrowth syndromes, and autism spectrum disorders ASD). We are interested in understanding how mutations in PTEN result in such varying clinical presentations. We predict that mutations alter PTEN function in different ways or degrees and this is the source of the clinical variability. To test this prediction, we will first use a new economical gene synthesis approach to make thousands of copies of the PTEN gene, each copy with a different mutation. This can be accomplished by assembling PTEN using a set of 11 small fragments as building blocks. We can preprogram different mutations into each fragment. These fragments are then assembled into the full-length PTEN using new bioengineering techniques. Importantly, these methods allow us to assemble thousands of PTEN copies in the same reaction making a mutation library. Then, we will test this mutation library using a 'humanized' yeast system. In this system, the normal yeast genes are replaced with the human analogs. However, an overactive human kinase is added that is toxic to the yeast when expressed resulting in death). The expression of the normal human PTEN suppresses toxicity through its phosphatase activity. Using this system, we can simultaneously test thousands of PTEN mutations for their ability to rescue the yeast. With this data we evaluate the genetic risks of ASD and other patients with novel mutations. It may also explain why PTEN mutations have such diverse clinical presentations. If successful, these methods can be adapted to study the effect of mutations in other genes important in neuropsychiatric and neurodevelopmental disorders.