Breathing Effect of (AB)n-type protein assemblies

[xyfzkd]

In (AB)n-type protein assemblies, the breathing effect refers to small, dynamic changes in the conformation of the protein due to thermal fluctuations or internal interactions. This effect can cause the alternating A and B subunits to shift slightly, leading to structural variability. In cryo-EM reconstructions, such movements can result in gaps or blurred regions because the averaging process struggles to capture these subtle fluctuations accurately.
If we think of the protein as a barrel, each unit is like a vertical plank on the side of the barrel, but the connections between the planks are not very rigid. As a result, the opening of the barrel may vary in size.

My case is…
Initially, when reconstructing all the particles, you can obtain clear density for one AB subunit, but for other AB subunits, the density is quite poor, even dispersed.
Using 3D classification reduces the dataset to 1/5 of the original size, leaving fewer than 100,000 particles. In my (AB)3 protein, I can clearly see two pairs of AB subunits, but the third pair remains blurry. In this case, the structure is insufficiently symmetric, so it is not suitable to add symmetry options for further reconstruction. Additionally, there aren’t enough particles to classify again to obtain those with better symmetry. My understanding is that there may be subtle differences in the diameter of each initial particle. How can I make the best use of all the particles to obtain one or more structures?

Thanks.

Best,
Yufeng

[olibclarke]

Hi Yufeng

You might want to read this:

Case Study: End-to-end processing of encapsulated ferritin (EMPIAR-10716)

We’re excited to be sharing a case study on the end-to-end processing of encapsulated ferritin in CryoSPARC! This case study demonstrates some advanced tools and processes that enable processing of a structure with unconventional symmetry and geometry characteristics. Topics covered include: Ab-Initio Reconstruction of high-symmetry datasets handling local symmetry and non-point-group symmetry use of Volume Alignment Tools and Align 3D Maps for custom geometry operations jobs such as 3D Classi…

Which addresses exploiting non-point group symmetry in a pseudo-symmetry expansion approach, which may be something worth trying for your data.

Cheers
Oli

[xyfzkd]

@olibclarke

non-point

I appreciate your kind advice. In the case you mentioned, a key point is resolving internal orientation diversity through group re-alignment, which is similar to what I did in my own case. I expanded the C3 symmetry to align the worst, modest, and best densities of the subunits. Local refinement of these three datasets helped correct small orientation deviations, and I was able to obtain a high-resolution density for one subunit while the other subunits remained smeared. However, I am wondering how I can achieve an entire density, as there might be a second high-resolution density for the full structure.

Thank YOU!