From: Arturo J. Morales <art@scripps.edu>
  To  : Dr A.J. Rowe <ajr@leicester.ac.uk>
  Date: Tue, 05 Jan 1999 11:24:29 -0800

Re: Interesting problem...

I have been using Svedberg to fit to MW....  I should have specified that :)

Using Svedberg, I get the following fit for the RNA alone:

S= 3.791 +/- 0.003
M= 24.42 +- 0.33
RMS residual 1.67E-2
using vbar=0.53

I've also used dc/dt to get s using g(s) vs s... I mixed approaches when
describing the data in my email (which might actually reflect the status of
my brain at this time...) in fact, the s for RNA was larger than 3 ...
around 3.4 from what I can estimate from some of the graphs upon closer
examination...

I just found out Beckman is installing new software on our machine
tommorow, so I'll try  SE and SV at higher speeds after the machine comes
back up :)

At 06:08 PM 1/5/99 +0000, you wrote:
>Art -
>
>There are some puzzling features about your data. As John Philo as
said, how 
>are
>you calculating M from s + vbar ?  On the assumption that all species
>under consideration are more or less globular (bit dodgy for tRNA) then
the 
>Squire Himmel equations can be used to estimate M values. From your data,
these com
>out as:
>
>RNA		s = 3.0S	vbar = 0.53	M = 14823	+/-978
>
>Protein		s = 2.0S	vbar = 0.7595	M = 21923	+/-1447
>
>So, whilst I agree that your M value for the protein does not differ
>significantly from that expected (24 kD), the RNA estimate is wildly out as
>compared to tRNA. And whilst correcting for asymmetry would be a step
in the
>right direction, its just not a big enough effect in this case.
>
>IN PRINCIPLE it ought to be possible to work out the stoichiometry from
s 
>values
>alone, since there are only a very small number of real possibilities
(1:1, 
>1:2,
>2:1, 2:2 ?). You just work out the s value for each, using an appropriately
>weighted vbar and the Squire-Himmel equations. And see which estimate fits.
>However, that is only valid if you get the right M for each individual
species,
>and you do not.
>
>A further complication. There is no plausible way in which a 2S species can
>combine with a 3S species to give a 2.5-2.8 S complex, even at 1:1. Well,
short
>of wildly improbably huge unfolding of a component on association. 
>

My thought is that as they bind and the vbar changes, the apparent s
changes too... is that possible? so that a direct comparison of the change
in s is not possible unless we take vbar into account.

Thanks again!  (to all who have helped me make sense... :)

Art


----------------------------------------------------------------------
     Arturo J. Morales    (RPI
'94)             art@scripps.edu
   Department of Biology - Massachusetts Institute of Technology &
  Skaggs Institute for Chemical Biology - Scripps Research Institute
                   http://schimmel.scripps.edu/~art