There is a wealth of servers, freely available online, which allow analysis of any protein sequence. RaptorX was recommended to me by Dr. Yufeng Tong as an excellent structure prediction server so I decided to give it a go with the huntingtin amino acid sequence. All of the output from this work can be seen on Zenodo.
In summary, this server threw up some rather different predictions for the huntingtin protein structure. Although the overall structure is dominated by features of HEAT repeats bundled into a solenoid structure as I have seen previously, the prediction suggests that the solenoid is continuous throughout the entire huntingtin structure instead of discrete domains.
Disordered or less structured regions are still visible in the characteristic regions of sequence (exon1, amino acids 400-600 and so forth) but they are extended beyond the alpha helical solenoid and do not interrupt the overall structure.
Raptor X models for different regions of the huntingtin protein sequence
This model does still tally with the experimental data; these loops would be proteolytically accessible and under denaturing gel conditions the resultant fragments would be individually resolved, even if they were still interacting with each other by non-covalent interactions.
Again, without further experimental data. it is difficult to read much more into these models although they I am increasing convinced that huntingtin is either a continuous solenoid or a string of solenoid domains.