Dendritic Spine Development in Dup15q and AS

Quick Overview

Les Loew presented a talk on dendritic spine development in Dup15q and AS, focusing on the integration of information to make predictions. The work was carried out by two graduate students, Judy Bloom and Olena Marchenko, along with collaborators. They used iPSC cells from Angelman and Dup15q patients, differentiated them into neurons and performed morphological analysis. The Dup15q cells were not highly branched, while the neurotypical cells had a reasonable level of branching. The Dup15q cells were more highly branched. The team also looked at dendritic protrusion density, finding a large number of spines in Dup15q cells. The team used a modelling software system called VirtualCell to build a comprehensive mathematical model of the ubiquitin pathway. The model showed that slight variations in UBE3A may aid in spine development. The team concluded that UBE3A may regulate actin dynamics in dendritic spines through the Rho GEF ephexin5.


Thanks to the organisers for putting this all on. It’s really stimulating to see all the different ways that these two disorders that we’re so interested in can come about. One of the themes of my talk is to try to think about how we can integrate or organise all this information together and actually do something to make some predictions.


This work was done largely by two graduate students, Judy Bloom and Olena Marchenko, and our collaborators, Stormy and Eric, and the people in their lab who have been working most directly with us, Noel and James. Michael Blinoff, a computer scientist, has developed a lot of the modelling approaches that we use.

Experiments and Modelling

Most of the talk will be on modelling, but I’ll show you some experiments that are motivating us. We’ve taken these iPSC cells, both from Angelman patients and from Dup15q patients, and we’ve grown them up, differentiated them into neurons, and did some morphological analysis.


The Dup15q cells are not very highly branched. The neurotypical cells have a reasonable level of branching at 30 weeks, and the Dup15q cells tend to be quite a bit more highly branched. The second kind of analysis that Olena in the lab did was to look at dendritic protrusion density.

Actin and Dendritic Protrusions

Actin is the key cytoskeletal element in dendritic protrusions. Actin and actin dynamics control dendritic protrusions and dendritic morphogenesis.

UBE3A Correlation

UBE3A is a key component of the system. When a dendritic protrusion finds a bouton on a neighbouring axon, it engages some receptors called EphB receptors. They release Ephexin5, which also gets phosphorylated in the process. Ephexin5 can be degraded in its phosphorylated form by UBE3A.

Modelling Approach

Over the last 20 years, my lab has developed a modelling software system called VirtualCell. It’s a software tool for modelling cell biology. We can solve ordinary differential equations, sort of kinetic systems. We can solve partial differential equations where the kinetic systems are associated with specific geometries. We can do stochastic simulations.


Human neurons from cultured AS and Dup15q stem cells develop respectively deficient and enhanced levels of dendritic branches and protrusions. Regulation of actin dynamics underlies the variation in dendritic morphogenesis. UBE3A may regulate actin dynamics in dendritic spines through the Rho GEF ephexin5. We can recapitulate at least the ubiquitin pathway in a virtual cell rule-based model. Virtual cell can serve as a platform for organising and integrating our knowledge of dendritic spine physiology.

Talk details

  • Title: Dendritic Spine Development in Dup15q and AS
  • Author(s): Les Loew
  • Author(s)’ affiliation: University of Connecticut
  • Publication date: 2016-09-06
  • Collection: 2016 ASF-Dup15q Scientific Symposium