Rg

Linear (OLIGOMER), and non-linear (MIXTURE) analysis, singular value decomposition (SVDPLOT), addition of missing fragments (BUNCH, CORAL), analysis of flexible systems (EOM/RANCH & GAJOE), flexible refinement of high-resolution models (SREFLEX)
Message
Author
hab44
Member
Posts: 10
Joined: 2016.12.24 22:05

Rg

#1 Post by hab44 » 2016.12.24 22:19

Hi,

I have a question about the radius of gyration.

I understand that the Rg value for a flexible protein or an elongated protein will be greater than that of a globular protein of the corresponding molecular weight. Is the Rg for a flexible protein greater than that of a globular protein because it is extended? Or is it because of the flexibility? I want to understand whether it is the extended nature of flexible proteins that gives them a higher Rg than a compact/well-structured protein of the same molecular weight. :?

Many thanks in advance for your help,

Heather

User avatar
AL
Administrator
Posts: 686
Joined: 2007.08.03 18:55
Location: EMBL Hamburg, Germany
Contact:

Re: Rg

#2 Post by AL » 2017.01.02 18:31

hab44 wrote:Is the Rg for a flexible protein greater than that of a globular protein because it is extended? Or is it because of the flexibility?
It's because it is more extended. Have a look at Fig. 2 in Kikhney & Svergun 2015 review - maybe it will help. Also check out Flory’s equation (6) two pages further.

hab44
Member
Posts: 10
Joined: 2016.12.24 22:05

Re: Rg

#3 Post by hab44 » 2017.01.11 19:29

Thank you! This helps a lot.

hab44
Member
Posts: 10
Joined: 2016.12.24 22:05

Re: Rg

#4 Post by hab44 » 2017.01.11 20:35

Hi Al,

I have another question about Rg. If there were two proteins with the same number of residues, but one was intrinsically disordered and the other completely rigid like a straight rod, which protein would have the higher Rg? As I understand it, the flexible protein would be sampling more space but the rigid protein would have a greater maximum length.

Once again thanks for your help,

Heather

ckerr
Active member
Posts: 86
Joined: 2015.03.25 09:03
Location: EMBL Hamburg

Re: Rg

#5 Post by ckerr » 2017.01.12 09:33

The straight rod would have higher Rg

hab44
Member
Posts: 10
Joined: 2016.12.24 22:05

Re: Rg

#6 Post by hab44 » 2017.01.12 13:33

Dear ckerr,

Thank you for the information. The reason that I ask is that I have a flexible protein (30 kDa) with Rg 52 A (Guinier derived). This flexible protein interacts with a globular protein (50 kDa) and the resulting binary complex has an Rg of 53 A (Guinier derived). The globular protein sits somewhere in the middle of the flexible protein, so not on one end and therefore not expected to contribute to Dmax. As the Rg values are so similar despite a difference in mass of 50 kDa, I am unsure about what is happening in this system (although I have considered different possibilities, I would like an expert's view).

So far I have used EOM to look at the distribution of Rg and Dmax values an ensemble of the flexible protein would have. From the EOM analysis, the mean Dmax is 180 A and range is 80 A to 270 A. I have also used DAMMIF to generate a molecular envelope of the binary complex. The P(r) function was generated in GNOM and Dmax was 225 A. One possibility is that the binding of the globular protein to the flexible protein induces some structural compaction in the latter, so reducing the extendedness of the complex and resulting in similar Rg despite increase in mass. But is this interpretation correct considering that the Dmax of the binary complex is greater than the mean Dmax of the flexible protein? What should I rely on more - EOM or DAMMIF - for this complex? The Dmax was difficult to pin down in GNOM for the binary complex, possibly due to the flexibility at both ends of the complex.

Any ideas about this would be much appreciated,

Heather

ckerr
Active member
Posts: 86
Joined: 2015.03.25 09:03
Location: EMBL Hamburg

Re: Rg

#7 Post by ckerr » 2017.01.12 15:39

If you put the globular protein in the middle of the flexible protein without affecting the structure of the flexible protein then you would effectively have a core-shell structure with a dense core (the globular protein plus the inner part of the flexible protein) and a less dense shell (outer part of the flexible protein), so the Rg would actually decrease relative to the flexible protein on its own.

However, this will not happen because (a) the flexible protein is excluded from the volume occupied by the globular protein, and (b) there must be some sort of binding interaction between the two proteins, otherwise a complex would not form.

Some possible interpretations (there are probably others I didn't think of):
1) The complex has significant lengths of flexible chain. This produces a 'halo' around the globular core, which contributes to Dmax but, because of its low density, has little effect on Rg
2) The complex is globular but prolate (e.g. like a short fat rod).

hab44
Member
Posts: 10
Joined: 2016.12.24 22:05

Re: Rg

#8 Post by hab44 » 2017.01.12 16:18

Thank you - I appreciate your insight.

In 1) what do you mean exactly by 'significant'?

In 2) do you think it is possible that the well-structured/globular protein confers some rigidity to the flexible protein? So the flexible protein becomes more structurally ordered upon interaction with the well-structured/globular protein but retains length/extendedness? One thing that I did not mention is that the flexible protein has a high density of basic residues and proline residues. The predicted Rg for the flexible protein is 45 A assuming that it is a typical IDP (from Flory's equation using R0 and v parameters in Tria et al., 2015). Yet the experimental Rg at 52 A is higher.

ckerr
Active member
Posts: 86
Joined: 2015.03.25 09:03
Location: EMBL Hamburg

Re: Rg

#9 Post by ckerr » 2017.01.12 18:55

1) Enough that it is detectable in the SAXS data

2) If the flexible protein binds in more than one place then that would convey some rigidity. To be honest I'm just guessing here so don't take my word as gospel.

Alex
Active member
Posts: 692
Joined: 2007.08.09 21:10
Location: Planet Earth

Re: Rg

#10 Post by Alex » 2017.01.13 12:34

I think the real question here is how to differentiate an extended rigid protein from flexible protein using SAXS?

I think in general one should have an idea if the protein is rather extended (e.g. has a rod shape). If not, one can check dimensionless Kratky plot with
references to well folded and natively unfolded protein.

Personally, i find it difficult to differentiate extended "rigid" structure from flexible one using SAXS invariants or Kratky plot. This is even more difficult if the structure can be somewhat extended and/or somewhat flexible..

hab44
Member
Posts: 10
Joined: 2016.12.24 22:05

Re: Rg

#11 Post by hab44 » 2017.01.13 12:58

Very little information about this flexible protein is available, but bioinformatic structure predictions suggest that a coiled coil region is present at one end, which may contribute to the extendedness of the flexible protein alone, and in complex with the globular protein. The dimensionless Kratky profile for the flexible protein supports high degree of flexibility whereas the binary complex is predominantly structured with some residual flexibility and is also extended. DLS supports an elongated shape for the binary complex.

For me, the most intriguing part of the data was the Rg similarity between the two particles, and I don't want to interpret this incorrectly.

The comments are very useful as I understand the system a bit more now - thanks very much. If anyone has any other thoughts, please let me know.

Alex
Active member
Posts: 692
Joined: 2007.08.09 21:10
Location: Planet Earth

Re: Rg

#12 Post by Alex » 2017.01.13 13:15

The predicted Rg for the flexible protein is 45 A assuming that it is a typical IDP (from Flory's equation using R0 and v parameters in Tria et al., 2015). Yet the experimental Rg at 52 A is higher.
there can be a concentration effect too. Did you extrapolate the data to infinite dilution?

The reason that I ask is that I have a flexible protein (30 kDa) with Rg 52 A (Guinier derived). This flexible protein interacts with a globular protein (50 kDa) and the resulting binary complex has an Rg of 53 A (Guinier derived).
if you look at natively unfolded proteins, like hTau40wt (MW ~44kDa) it has Rg of 55-60 A. Rg for BSA (MW 66) is about 30 A. Of course these are just some references, but it seems like your complex becomes somewhat more rigid. Given that, why don't you try modeling complex with CORAL and EOM, although i believe the latter (though possible) might be not straightforward.

Alex

hab44
Member
Posts: 10
Joined: 2016.12.24 22:05

Re: Rg

#13 Post by hab44 » 2017.01.13 13:33

The data was collected with HPLC directly before the SAXS measurement, so I plotted Rg across the eluting peak for both samples. The Rg was consistent across both peaks. For each sample, frames from the centre of the peak where averaged to generated the scattering intensity curve. Data on the buffer was collect before sample elution and subtracted from the sample. The Guinier plot does not show any upward or downward turn at the start of the curve.

I used EOM for the flexible protein, but was unsure about how best to do this for the binary complex. I gave is a try by inserting the sequence of the globular protein directly in the middle of it's predicted binding site in the flexible protein. I don't think that I did this right though, so didn't rely on the results...

Alex
Active member
Posts: 692
Joined: 2007.08.09 21:10
Location: Planet Earth

Re: Rg

#14 Post by Alex » 2017.01.13 14:56

used EOM for the flexible protein, but was unsure about how best to do this for the binary complex. I gave is a try by inserting the sequence of the globular protein directly in the middle of it's predicted binding site in the flexible protein. I don't think that I did this right though, so didn't rely on the results...
is there any partial structure for the flexible protein available? I mean do you have a "minimal" interface where you would have rigid protein and partial structure of a flexible one?

hab44
Member
Posts: 10
Joined: 2016.12.24 22:05

Re: Rg

#15 Post by hab44 » 2017.01.13 15:17

Unfortunately there is no high resolution data for the flexible protein alone or even a small segment bound to the globular protein.

The flexible protein is 248 residues long. In the literature the site on the flexible protein where the globular protein will bind has been narrowed down to residues 96 to 160. We also have Hydrogen/deuterium exchange mass spec data which maps the interaction site of the flexible protein on the globular protein. Can this information be used as constraints in CORAL?

Post Reply