[RASMB] protein polymer analysis

John Philo jphilo at mailway.com
Tue Jul 29 08:22:55 PDT 2014


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




More information about the RASMB mailing list