Today I’ll be looking at the 3D structure of my pet gene ASPA. When looking through the NCBI structure database I decided on using the entry with the PDB ID of 2O53. It was one of the most recent 3D models and appeared to have to no alterations. This protein came as a dimer but I hid one of the regions to better examine my target area. On my last post I mentioned the E285A mutation that has been found to cause Canavan disease and I compared the wild type protein to the E285A variant with various structural modeling algorithms. Today I’ll be looking at the 285th AA in the actual 3D model.
Below you can see the space filling model. While hard to see the highlighted region is in a small concave on the 3D model. This is believed to be where the catalytic domain is.
In this next picture is a little bit closer examination of the target spot. From here you can see that the 285 spot is part of a turn between two beta sheets.
When actually examining the 3D structure I found quite a few of the predictions the algorithms gave me in my last post were incorrect or slightly off. First off the 3D models shows that the 285 region is actually a simple turn as you can see above. The Protean 3D algorithm guessed that area would be an alpha or beta region and not a turn. Protean 3D also predicted that the 285 area would be hydrophobic and the area directly after would be hydrophilic. As you can see from the picture below the 285 region is actually hydrophilic. The predictions weren’t totally wrong because the model does show a hydrophobic region followed by a hydrophilic region just a few AAs down the protein.
The charge region was also a little off as you can see in the picture below. The 285 spot is negative as predicted but Protean 3D also predicted that a few AAs on each side would be negative. But in the 3D model it shows that the AAs around the 285 spot are neutral.
Overall I would say that the Protean 3D algorithms had some issues predicting what this protein would actually look like. The predictions were definitely helpful when seeing what the E285A variant would do but the actual 3D model has quite a few differences.
michelconn 