Dear Ose,
I think we have met similar problems. My protein is relatively small, ~17kD, but it could form aggregates larger than 100nm. A classic Protein Science paper in 2003 by DR Smith titled "Crystal structures of fusion proteins with large-affinity tags" offered me great help. With MBP tag and optimized linker my protein became significantly less aggregated (~10nm).
However, this is still far from a guarantee for success since it is still quite difficult to obtain crystals, although Jovine's group reported a recent success about ZP-N.
By the way, can you observe a monomer peak during ultracentrifuge? In terms of DLS, how about the polydispensity? HPV E6 also aggregates to similar size, but the granules were relatively homogeneous. More importantly, it was observed that in cell the protein could still form such large aggregates. Maybe for some proteins, they were just meant to be like this. And based on my very limited knowledge, most structures of such proteins remained as the "higher-hanging fruit".
Last but not the least, J Jancarik, et al. designed a elegant protocol (2004，Acta Cryst. D60. 1670-1673.) to find the optimized buffer that could increase the solubiltiy and homogeneity, whichmight be helpful too.
Good luck!
Xuan Yang
2009/8/27 ose toyoyuki
<[log in to unmask]>
Dear all,
This is a question on how to cope with the protein that seems to form
massive aggregates in solution but enzymatically active.
I'm working on a protein whose molecular weight is around 70kDa and can be
divided into two domains (say A and B domains). We expressed this protein in
E.coli fused with GST and purified using some chromatography. The GST
affinity chromatography works well and proteinase digestion to remove the
tag does wonders too. The purified protein was confirmed to be active
enough, we can detect both activities from these two domains. But the
retention time from the gel filtration clearly shows it is awfully
aggregated (comes out at the void region). DLS measurement indicates the
averaged diameter is around 45 nm, which I feel is a bit too long.
Analytical ultracentrifuge result implies that the distribution of the
molecular species is wide, some portion got precipitated with 1K rcf (means
the molecular weight is more than 5MDa) and the rest is ranging from 1MD to
5MDa with a peak at 1MDa.
I made new two constructs covering the A and B domains respectively, both of
them are active again, but only the A domain has got the same symptom as the
intact protein. The B domain seems to exist as a monomer in solution.
Here come my questions, (I) How can I interpret this phenomenon? (II) Is
there anything we can try to change the situation? (III) Does it make sense
to try crystallization? (probably not).(IV) Has anyone got such experience?
I tried the methylation on lysine side chains, I also tried the buffer with
0.2M arginine or 10% glycerol but the all the results just seem hopeless.
The protein before the proteinase treatment also comes out at the void
region from the gel filtration.
cheers
toyoyuki
--?
Toyoyuki Ose
[log in to unmask]
Graduate School of Life Science
Hokkaido University
N21W11 Kita-ku, Sapporo
001-0021 Japan
