Influence of iron-heme on scattering data?

Interactive and automated data processing tools (PRIMUS, GNOM, AUTORG).
Scattering from simple bodies (BODIES), peak analysis (PEAK), data plotting (SASPLOT) etc.
Post Reply
Message
Author
Yann
Active member
Posts: 44
Joined: 2010.11.03 11:31
Location: Vrije Universiteit Brussel

Influence of iron-heme on scattering data?

#1 Post by Yann » 2014.03.06 11:55

Hi all,

I am working with the holo and apo form of a cytochrome c peroxidase. The MM of the apo form is 34.4 kDa. The presence of the iron-heme in the protein increases the MM to 35.1 kDa.

Both data sets were collected in batch mode. There were some small concentration-dependent effects, but after taking care of that these are stats.

Rg, apo = 20.27 A
Rg, holo = 20.36 A

I(0) apo = 6.094
I(0) holo = 7.734

What troubles me here is that the Dmax are really different, which is not what we expected. Based on other data (NMR, etc) we believe the apo form has a larger volume in solution, and then 'compacts' upon binding of the iron-heme ligand (this is also what we observe when we plot the Porod-Debye plot, I(q) q4 vs q4).

However, when performing the p(r) analysis we see for the apo form that the Dmax is around 55A (nice bell-shaped curve) while for the holo its around 70A (bell-shaped curve with a tail). For the holo it looks like there is some 'aggregation' (which does not make sense based on the quality of the Guinier regions and the overall scattering data) or like there is some extension (does not fit with the NMR solution structure). The ab initio models always show a nice 'ball' with a tail.

Now here is my question. Is it possible that the presence of the iron-heme introduces an 'aggregation artefact' in the data? Intuitively, I can imagine that the presence of an iron-heme greatly increases the scattering power of the protein without actually contributing much to the MM of the protein. Could PRIMUS mistakingly see this as a protein sample with some 'aggregation' instead of a globular protein containing an entity that increases the scattering power of the sample but not the MM?

If so, how do you process this correctly? I've provided the p(r) curves in attachment.

Does my question make any sense or am I overlooking something?

Looking forward to hearing your thoughts on the matter.

Thanks,

Yann
Attachments
Holo.jpg
Holo.jpg (67.9 KiB) Viewed 4681 times
Apo.jpg
Apo.jpg (63.05 KiB) Viewed 4681 times

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

Re: Influence of iron-heme on scattering data?

#2 Post by Alex » 2014.03.06 13:35

Hey Yann
Attaching the actual extrapolated data files would help.
Now here is my question. Is it possible that the presence of the iron-heme introduces an 'aggregation artefact' in the data? Intuitively, I can imagine that the presence of an iron-heme greatly increases the scattering power of the protein without actually contributing much to the MM of the protein. Could PRIMUS mistakingly see this as a protein sample with some 'aggregation' instead of a globular protein containing an entity that increases the scattering power of the sample but not the MM?
It should not introduce 'aggregation artefact'. The only problem I had when doing measurements on protein-iron systems was due to presence of unbound iron either in buffer or sample, which I don't believe is the case with your data. From what I can see, iron-heme increases your I0, but not that "greatly". It is you and not PRIMUS who can see things like that - so PRIMUS is not to blame here.

Cheers,
Alex

Yann
Active member
Posts: 44
Joined: 2010.11.03 11:31
Location: Vrije Universiteit Brussel

Re: Influence of iron-heme on scattering data?

#3 Post by Yann » 2014.03.06 13:59

Hi Alex,

wasn't really trying to 'blame' PRIMUS here. Maybe I formulated my concern the wrong way, mea culpa.

From your post I conclude that the tail I see in the p(r) function for Holo is real then?

We do have a concern that the iron-heme might 'leak' out from the protein as the iron-heme is NOT covalently attached in this case. We've noticed this kind of leakage before. So, this might then be the 'problem'.

How would you treat this the best way then? Add constant subtraction when performing ab initio modelling?

I've attached both curves.

Thanks again,

Yann
Attachments
Holo.dat
(31.95 KiB) Downloaded 178 times
Apo_Alex.dat
(31.95 KiB) Downloaded 156 times

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

Re: Influence of iron-heme on scattering data?

#4 Post by Alex » 2014.03.06 15:33

From your post I conclude that the tail I see in the p(r) function for Holo is real then?
It looks real to me. The only concert I have is that if you look at the fit of back calculated theoretical
curve to the experimental data (when calculating P of R), the low angle points do not fit well. So, check the way you merge/extrapolate your data.
We do have a concern that the iron-heme might 'leak' out from the protein as the iron-heme is NOT covalently attached in this case. We've noticed this kind of leakage before. So, this might then be the 'problem'.
is not it the data collected at Soleil with inline HPLC column?
i don't think the tail comes from "leakage" as it shows the probability of very long distances which
can't be from iron-heme.
How would you treat this the best way then? Add constant subtraction when performing ab initio modelling?
I would treat it as normal SAXS data: observe concentration influence, extrapolate if necessary and
calculate invariants to see whether they match values you are expecting. From what I see, I would not exclude that holo and apo forms are in different oligomer states actually.

HTH,
Alex

Yann
Active member
Posts: 44
Joined: 2010.11.03 11:31
Location: Vrije Universiteit Brussel

Re: Influence of iron-heme on scattering data?

#5 Post by Yann » 2014.03.06 18:36

Hi Alex,

no, the data was collected at the ESRF in Grenoble in batch mode.

Well, what could then explain the 'aggregation' is that the iron-heme does decrease the solubility of the protein, so we probably have the formation of some higher-order oligomers as you suggest.

Maybe a bit of a naive question: how could you tell the low-angle fitting was not good? In my version of PRIMUS (I work with a Mac) I get no residuals during p(r) analysis. Do you just plot the intensities (exp vs reg) from the .out file and compare?

Molecular mass analysis (Fischer, Qr method, etc) all give decent estimates for the MM and the extrapolated scattering curves fit very well with the crystal structures (chi2 = 1.2).

Thanks,

Yann

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

Re: Influence of iron-heme on scattering data?

#6 Post by Alex » 2014.03.06 19:22

Maybe a bit of a naive question: how could you tell the low-angle fitting was not good? In my version of PRIMUS (I work with a Mac) I get no residuals during p(r) analysis. Do you just plot the intensities (exp vs reg) from the .out file and compare?
I use Mac too. In P(R) you posted earlier - when you are in Primus just go to the left part displaying the fit of backcalculated curve to the experimental data and zoom in to the low angle part of the fit (i think you should hold mouse button and pull the cursor to do so).
Molecular mass analysis (Fischer, Qr method, etc) all give decent estimates for the MM and the extrapolated scattering curves fit very well with the crystal structures (chi2 = 1.2).
So, do you have similar oligomer states in both cases?

HTH,
Alex

Yann
Active member
Posts: 44
Joined: 2010.11.03 11:31
Location: Vrije Universiteit Brussel

Re: Influence of iron-heme on scattering data?

#7 Post by Yann » 2014.03.06 20:46

Ah ok, that simple :). Didn't know I could zoom.

Nope, the apo form does not seem to do that.

We do know that the holo can only interact with its partner when it has the iron-heme. When it is apo, it does not bind its partner. Maybe there is some self-self interaction once the protein becomes holo, which could explain the oligomerization...

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

Re: Influence of iron-heme on scattering data?

#8 Post by Alex » 2014.03.06 21:10

We do know that the holo can only interact with its partner when it has the iron-heme. When it is apo, it does not bind its partner. Maybe there is some self-self interaction once the protein becomes holo, which could explain the oligomerization...
Just calculate MW from SAXS curves using different ways. I am not sure, but I think one of them is a homodimer.

Yann
Active member
Posts: 44
Joined: 2010.11.03 11:31
Location: Vrije Universiteit Brussel

Re: Influence of iron-heme on scattering data?

#9 Post by Yann » 2014.03.06 23:30

Hi Alex,

I did that and always get around 34kDa - 35kDa which are the expected molecular masses for monomers.

Maybe some aspecific oligomerization in the case of the holo protein.

Thanks again,

Yann

Post Reply