Select Page

Goldilocks and the UBE3A: Finding the sweet spot in UBE3A gene therapy


(Yi et al.)

UBE3A mutations affecting its activity are related to autism and Angelman Syndrome

Summary by: Kristen Nordham

Edited by: Barbara O’Brien, Ph.D.

Original article: Jason J. Yi, et al. 2015. An autism-linked mutation disables phosphorylation control of UBE3A. Cell 162(4): 795–807. doi:  10.1016/j.cell.2015.06.045

Human genes are capitalized and italicized: UBE3A

Mouse genes are not capitalize but are italicized: Ube3a

Proteins are capitalized and not italicized: UBE3A


Increased UBE3A is related to Autism while decreases leads to  Angelman syndrome


Angelman Syndrome (AS) and autism are both neurological developmental disorders associated with nervous system function. A common genetic connection between AS and autism is the UBE3A gene. The UBE3A gene makes an enzyme called a ubiquitin ligase, and the function of this enzyme is to put a tag on proteins that are past their prime. Proteins tagged by ubiquitin are sent to the cell’s recycle center (the proteasome) for destruction.


We know that the deletion or mutation of maternal UBE3A causes AS. And having too many copies of maternal UBE3A is one of the most common known genetic causes of autism, such as in Dup15. Thus, since too little UBE3A leads to AS and too much UBE3A is connected with autism, it seems likely that precise control the amount of UBE3A protein, and how active it is, is necessary for normal brain development. Previous research suggested that the UBE3A protein tags itself for destruction, as well as tagging all the other proteins it targets. A very self-critical protein. So most scientists thought that abnormal UBE3A would disrupt its own self-regulation, and might cause UBE3A levels to vary out of control, both too high and too low.


However, the authors of today’s featured article from the Zylka Lab of the University of North Carolina considered another idea: If there were too much UBE3A protein, and it is targeting itself for destruction, it might completely self-destruct! Over-abundance could lead to total elimination of the protein. In other words, abnormal or excess UBE3A might get stuck in “destruction mode” and degrade itself more than expected, leading to decreased levels or a complete loss of UBE3A.


UBE3A mutations reduce UBE3A in three different ways

UBE3A marks cellular waste to be degraded, and if UBE3A activity is abnormal, this trash can pile up!


To test this theory, the researchers put the human UBE3A gene into human embryonic kidney cells in a petri dish. This is a common way to study cell development in neuroscience research. Then, they mutated the UBE3A gene to investigate what these mutations would do to the function of the UBE3A protein. Normally, UBE3A targets proteins for destruction by adding a small ubiquitin protein, which acts like a tag and marks the protein as “trash” to be taken apart and discarded so that it doesn’t take up precious real estate in the cell. When this activity is disrupted, cell machinery cannot properly identify this unmarked cellular matter, so it doesn’t get destroyed.


The researchers found three mechanisms by which missense mutations inactivated UBE3A, and these mutations have also been seen in human patients. Some of the mutations inactivated UBE3A by destabilizing the UBE3A protein. Others caused UBE3A to increase its own self-destruction! The rest disabled UBE3A so that it failed to target proteins for degradation with ubiquitin tags.  


The on-switch turns UBE3A off!

While phosphorylation often causes a protein to switch to its “active” state, UBE3A’s self-degradation is slowed by phosphorylation.


There are other ubiquitin ligases that cause diseases, and many of them are regulated by phosphorylation, which is the process by which a phosphate group is added to a protein. Proteins are made of amino acids, including threonine and alanine. Generally, when a protein is phosphorylated, this is a type of “on-switch”, putting the protein into an active “on” state, but sometimes phosphorylation can do the opposite, and inhibit a protein. The authors of today’s article wanted to know whether UBE3A activity could also be controlled by phosphorylation, like other ubiquitin ligases. They noticed that several of the gene mutations were located near each other. By looking into the center of this hotspot region of UBE3A, they found a specific site which is normally phosphorylated by an enzyme called protein kinase A (PKA). PKA does its job on UBE3A by phosphorylating a specific amino acid, the 485th threonine on the UBE3A protein (T485).


Wondering if this T485 site is really the on/off switch for  self-ubiquitination, the researchers changed the spot so that it could not be phosphorylated, by inserting another amino acid, alanine (A), which cannot be phosphorylated, in place of the threonine (T). This version was called T485A because an alanine (A) was traded for the threonine (T) at this location.


The researchers found that phosphorylation at T485 serves as an “off-switch”, inhibiting the self-ubiquitination of UBE3A, which slows  its degradation and increase UBE3A levels. Dephosphorylating (removing a phosphate group) at the T485 site, is like flipping the on-switch, increasing activity by UBE3A and speeding up  its own self-destruction. Other sites on UBE3A could be phosphorylated, but only T485 changed the amount of  UBE3A protein.


To see if what they had found in the cell cultures was also true in human beings, researchers collected DNA samples from children with autism and children with AS caused by UBE3A mutations, and all of their parents. They confirmed that alanine had been swapped with threonine at location 485, but only in the children with autism, not in the children with AS. This confirmed that the change of one amino acid at location 485 on UBE3A causes human disorders and their results suggest that increased UBE3A activity is what contributes to autism pathology.


Because PKA activates the on/off switch for the self-destruction of UBE3A, the authors considered the possibility of changing UBE3A levels by changing the amount of PKA. They found that it does work like this. Turning up PKA activity turns down self-degradation of UBE3A, most likely by extra phosphorylation of T485.


After concluding that reduced phosphorylation of UBE3A leads to autism, the researchers looked for a mechanism. How would over-phosphorylation or under-phosphorylation of UBE3A cause harm to neurons?


Turning UBE3A off impairs development of dendritic spines

(NeuwriteSD) Dendrites, or small growths off of neurons and receive messages from neighbor cells, develop abnormally when UBE3A activity is irregular, as in AS.


Dendritic spines are small projections that stick out from neurons to receive input from other neurons. If either the density of these spines is changed, or even if their shape changes,  , this communication between neurons doesn’t work properly. There are drugs that activate or inhibit PKA activity, and it is known that chronic inhibition of PKA leads to an increase in the density of dendritic spines. The Zylka Lab added these drugs into their cell cultures and found that, sure enough, the drugs that inhibited PKA also changed the activity of UBE3A. This was interesting because it links PKA and UBE3A activity, and hopefully means that there are more parts of the disease we could target in hopes for treatments. They also found that when they added extra UBE3A gene expression into wild-type cells, they developed more densely located spines compared to the control. But, when they added the PKA-inhibiting drugs into those cells, the spines developed normally. This experiment in cell cultures suggested that abnormal PKA and abnormal UBE3A could change brain development. But, would this still be true in animal models?  


To find out, the researchers genetically engineered mice to have mutations on the UBE3A gene so they could study whether UBE3A activity could be regulated by phosphorylation in mammals like in the cell cultures. They mutated the same T485 site that they had studied in the cells, the site where PKA phosphorylates UBE3A. While the mouse pups were still in utero, the researchers injected their brains with the mutated T485A version, which cannot be phosphorylated. They also injected some of the mice with an extra amount of the normal version of UBE3A, so that they would make too much UBE3A. At the same time, they injected a fluorescent tag into the mouse brains. The fluorescent tag glows when exposed to a certain light under a microscope so that the researchers could see where the mutated UBE3A had settled in the brain tissue.


A month after birth, the researchers looked at how the mice brains had developed. The Zylka Lab found that the mice which were overexpressing wild-type UBE3A showed only a small increase in the density of dendritic spines, while those with the mutated hyperactive T485A showed a large increase (54%) in dendritic spine density compared to controls. And, not only was the density of spines changed, the shape was also affected. In both the mice with too much UBE3A and the mice with mutated T485A, there was a decrease in the widths of the dendritic heads compared to controls. This decrease in width could affect how much signal the neuron is able to receive from other neurons, and that change could disrupt communication between neurons.


Next, the researchers wanted to see if a phosphorylated version of T485 would cause or prevent the abnormalities in dendritic spines. To do that, they replaced the threonine with glutamic acid (E), which mimics a phosphorylated threonine. Mutants with this mutant T485E showed normal spine density and shape compared to controls. This result confirms that the change in spine density and shape discussed above was caused by the missing phosphorylation at location 485.


In summary, the expression of overly active versions of UBE3A (hyperactive wildtype UBE3A and mutant T485A) caused the development of excess dendritic spines and this lasted into adulthood. The mutant which mimicked the inactive phosphorylated version of UBE3A, T485E), did not change the number or shape of spines. This shows that unphosphorylated versions of UBE3A caused abnormal neuronal spines while a phosphorylated version doesn’t! This points to the possibility that the phosphorylation of UBE3A by PKA controls proper neuronal development.




When UBE3A activity is abnormal, it can lead to both AS and autism, and even some cancers are related to abnormal UBE3A activity. Finding a way to control the activity of UBE3A with a drug could provide breakthroughs in treatments. These experiments by the Zylka Lab indicate that it could be possible to reduce UBE3A activity in individuals with duplicate or triplicate copies of the gene by using a drug which could stimulate PKA activity and tamp down UBE3A activity.


Specifically, they found that the T485 site in UBE3A is a hotspot for missense mutations which cause disease. Phosphorylation of this site has a master switch ability to affect UBE3A activity, this knowledge could lead to next steps in developing treatments. These findings also demonstrate that gene therapy to activate paternal UBE3A or increase UBE3A activity could help people with AS, but the levels of UBE3A will need to be carefully monitored, so it doesn’t get excessive and self-destruct. With UBE3A, nothing is ever simple, but this paper helps explain a lot.   

%d bloggers like this: