[RASMB] protein polymer analysis

[Tina Daviter]

Dear John,

Thanks a lot for your response which was very helpful in pointing out 
the limitations of each technique. It is one thing thinking about the 
theory, but practice is different. The real question we have is about 
potentially different polymerisation mechanisms of wt/mutants as the 
ladders look slightly, but reproducibly different. The question is 
whether secondary conformations of same-size oligomers are more 
populated in mutants. We wouldn't expect a direct answer from any 
technique, but I wanted to get experienced feedback on what to expect, 
which differences may be captured in AUC/SEC-MALS and if these might be 
interpretable with or without complementary data.

We can home in on short oligomers to reduce species numbers, but even 
then the % signal per species is admittedly small, certainly for less 
populated conformations. I expect getting reproducible, detailed 
information for several species was wishful thinking.

Thanks again!
Best wishes
Tina






On 29/07/2014 16:22, John Philo wrote:
> Tina, first, I think the group could better address which approach is better
> (SV versus SEC-MALS) if you could more clearly state exactly what you hope
> to learn about these mixtures.
>
> Regarding SEC-MALS, the molar mass is calculated for each slice of the
> chromatogram (usually several points per second), not averaged over each
> peak, and the calculations for each injection are independent of the others.
> But the molar mass is the weight-average of all species that are present in
> the detector at that moment. Thus if the resolution of your different
> oligomers is poor you will not get the molar mass of a single species. I
> think it is doubtful you could good separation of tetramer from pentamer
> from hexamer etc., and no single SEC column is going to cover the full range
> from monomer to 20-mer. Probably the best you could do to resolve the
> oligomers below ~decamer would be to run two Tosoh TSK-G3000SWXL columns in
> series.
>
> In principle the MALS detector can give you some conformation information
> from the 'rms radius' ('radius of gyration') values for each slice that are
> determined from the angular dependence of the light scattering intensity.
> However those values are not very well determined for species as small as
> your monomer. But if the chromatography is working correctly the elution
> positions are of course related to the hydrodynamic size and thus to
> conformation (but that info has nothing to do with the MALS detector).
>
> Overall the weakness of SEC-MALS is that it only works well if the
> chromatography works well, and too often the protein wants to interact with
> the column matrix which can seriously affect the resolution.
>
> Your AUC questions are complex and I can't take the time to address them in
> detail. But my general comment is that to my knowledge no one has
> demonstrated the ability from a single SV run to reliably get both
> sedimentation coefficient and molar mass information (and thus shape
> information) for more than ~5 species, and even to do that they all have to
> be in roughly equal proportions. When you are talking about species
> representing only 5-10% of the total signal, there is very little
> information about their diffusion present in the raw data, and to get the
> molar mass with enough precision to distinguish between say heptamer and
> octamer with reasonable confidence, in the presence of other oligomers, is
> in my opinion likely impossible.
>
> Having said that, if you can assign each oligomer based on its sedimentation
> coefficient (assuming you have a full ladder, and each oligomer has only one
> shape) then yes you could quantify the different oligomers and deduce their
> f/f0 ratios from their sedimentation coefficients.
>
> John
>
> -----Original Message-----
> From: RASMB [mailto:rasmb-bounces at list.rasmb.org] On Behalf Of Tina Daviter
> Sent: Tuesday, July 29, 2014 7:30 AM
> To: rasmb at list.rasmb.org
> Subject: [RASMB] protein polymer analysis
>
> Dear all,
>
> I'd like some opinions on what to get out of running AUC and SEC-MALS from
> samples that are protein mono/oligo/polymer mixes with different
> distributions (monomer up to about 20mers). The monomer size is 52 kDa, and
> samples can be prepared the polymer ladders of which look different on
> native gel. Also, some species have visible double bands, suggesting
> different shapes. I attach a picture. Polymers are extremely stable and
> don't re-distribute upon dilution, yet, SEC does not result in separated
> peaks I am told.
>
> Questions are, for AUC,
> - Can we expect peak separation in AUC?
> - Can we get reliable shape information for each peak? Which model to use
> for fitting, straight into c(s,ff0)?
> - Would running the monomer alone help for data analysis of the polymer?
> (How?)
> - What speed to run at if we want to focus on, say, tetramer to decamer
> size?
>
> Regarding SEC-MALS, it's inviting as it's so much quicker, but not sure
> about information content:
> - Would slicing a peak in exactly the same way for different samples still
> give information on the different size distributions, even though it's
> averaged?
> - I expect no shape information to be had?
>
> Any input/experiences regarding this would be most welcome.
>
> Thanks a lot in advance!
> Best wishes
> Tina
>
>
> --
> Dr. Tina Daviter
> ISMB Biophysics Centre
> Birkbeck, Univ. of London
> London, UK
>
> T 020 70790749
> E t.daviter at mail.cryst.bbk.ac.uk
> W http://biophysics.ismb.lon.ac.uk
>

-- 
Dr. Tina Daviter
ISMB Biophysics Centre
Birkbeck, Univ. of London
London, UK

T 020 70790749
E t.daviter at mail.cryst.bbk.ac.uk
W http://biophysics.ismb.lon.ac.uk