[RASMB] Frict ratios and sedfit

[Arthur Rowe]

Samantha

I think Karl's comments are very relevant. And they bring up a point which
is often overlooked when doing SV analysis. Which is this: a 'single' peak
seen by c(s) - or any other - analysis, reflecting the presence of a
'single' sedimenting boundary, does not of necessity imply the presence of a
single physical species. A monomer-dimer equilibrium also gives rise to a
single (and in practice only slightly asymmetric)
boundary/c(s)peak/whatever. {Gilbert & Jenkins, long ago}. So where
self-interaction is suspected - and even for choice when it isn't - it is
sound practice to do your SV analysis at 3 different cell loading
concentrations. If the measured s goes up with concentration, rather than
down, you have self association, for certain. To attempt to use a simple
c(M) fit to such a boundary is meaningless.

So - in answer to your original query, I do not think it is feasible, from a
single SV  experiment, to sort out your problem. It is, however, quite
feasible, as Karl suggests, that you have a tetramer-octomer in reversible
equilibrium. The higher c values which (I guess) you use for XRS would drive
the association towards octomer - and additionally all scattering methods
'see' large species better than small ones.

With Sedfit, as with all software, it is important not to let numbers
spilling out deprive you of your scientific judgement. If floating the
meniscus, for example, results in the meniscus being 'found' to be in a
position which is patently absurd, then do not believe it. And for sure, as
Karl says, do not swallow the idea of any sort of globular protein having a
frictional ratio >2. {OK - my old favourite protein, myosin, has an f/f0 >
3! But that's another story, and another field},

Kind regards

Arthur


--
*******************************************************
Arthur J Rowe
Professor of Biomolecular Technology
NCMH Business Centre
University of Nottingham
School of Biosciences
Sutton Bonington
Leicestershire LE12 5RD   UK

Tel:        +44 (0)115 951 6156
           +44 (0)116 271 4502
Fax:        +44 (0)115 951 6157
email:      arthur.rowe at nottingham.ac.uk
Web:        www.nottingham.ac.uk/ncmh/business
*******************************************************



Samantha,

Were the scattering studies done at much higher concentrations than the AUC
studies?  Were the solution conditions used in both experiments identical?

My understanding of the c(s) method is that if you have a dominant peak, the
transformation to c(M) from the optimized c(s) fit will give a good estimate
of the molecular weight.  The reason for this is that the wt avg f/fo ratio
is mostly reporting on the dominant peak.  This can change if some of the
smaller peaks have significantly different individual f/fo values.

I am not too sure how you did your c(M) modeling.  Did you float f/f0 using
the c(M) model?  In your original c(s) modeling, are you sure you had
reached the minimum reduced chi square?

I have recently had good success with the c(s,f/f0) model.  This model is a
lot like (if not exactly like?) doing a series of c(s) fits at different
constrained values of f/f0, then applying the regularization in both the s
and f/fo dimensions.  If your data consists of a predominant peak and is not
too noisy, both the s dimension and the f/f0 dimension will be well defined,
which allows you to estimate the average molecular weight of the main peak.
In this case, this estimate is not affected by the other peaks.  This is
described in Brown and Schuck Biophysics, 2006.

Also, I think a f/f0 of 2.4 is absurdly high --  on the order of an unfolded
protein.


Good luck,

Karl Maluf

________________________________

From: rasmb-bounces at server1.bbri.org on behalf of Samantha Jones (PRION)
Sent: Tue 10/31/2006 11:48 AM
To: rasmb at server1.bbri.org
Subject: [RASMB] Frict ratios and sedfit


Hello
I recently analyzed some data using c(S) distribution model on Sedfit. The
distribution showed a few peaks, with the main peak being at 4S. I floated
the frictional ratio and came back with approx 1.5. I then converted this to
a c(M) distribution with the main peak now showing approx 65,000 Daltons.
Knowing our monomer weight as 16,400 Daltons I concluded this was a
tetramer.
At the same time the same protein sample was analyzed using synchrotron X
ray scattering. Their results concluded an octomer, a two ring structure.
With this knowledge I went back to the data and reanalyzed using Sedfit,
c(s) and c(M) again. Using the calculator function on Sedfit I calculated
that for 4S with a frictional ratio 1.5 this would be roughly the tetramer
weight, however, a frictional ratio of 2.4 for 4s gives a molecular weight
equal to the octomer. When I entered this into the parameters of the c(M)
model the rsmd values show the f/f0 of 2.4 gives a better fit, 0.07 to 0.119
for f/f0 of 1.5.
My question is how I could have got this answer without the knowledge of the
scattering data? Is the use of the calculator function to fit data
recommended? Especially in a sample with three peaks?

Many thanks

Samantha 

Samantha Jones
MRC Prion Unit
Institute of Neurology
Queen Square
London
WC1N 3BG 

###########################################

This message has been scanned by F-Secure Anti-Virus for Microsoft Exchange.
For more information, connect to http://www.f-secure.com/
_______________________________________________
RASMB mailing list
RASMB at rasmb.bbri.org
http://rasmb.bbri.org/mailman/listinfo/rasmb



This message has been checked for viruses but the contents of an attachment
may still contain software viruses, which could damage your computer system:
you are advised to perform your own checks. Email communications with the
University of Nottingham may be monitored as permitted by UK legislation.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.rasmb.org/pipermail/rasmb-rasmb.org/attachments/20061102/16359dd3/attachment.htm>