[RASMB] fringe deviation

[Arthur Rowe]

Jack -

I agree very much with your argument that one should always do velocity
first. It is a great way of spotting what species are there, and you can get
a Kd value out. So you may then be in a good position to design your
equilibrium work (as you say, if not too much junk is present!).

Just one little extra point. The hydrodynamic non-ideality is not usually a
limiting problem. In 100 mM salt the monomer s-c behaviour is predictable,
whilst under conditions (e.g. very low salt) where simple theory will not
work, then you can still get on pretty well with an empirical s-c curve for
the monomer - to which one can normally get access. But strangely enough not
only do you not have to work above Kd, it is actually better in practice if
you work below Kd. Even a log or two below. This derives from the fact that
one normally knows how the monomer species behaves, and if one is nearer to
monomer than dimer then the fitting algorithm is rather insensitive to the
exact values which you assume for s(0)-dimer and ks-dimer.

Regards to all

Arthur


This burst of emails on fringe deviations has touched on a comparison of
SE vs SV, and the ease of including or avoiding hydrodynamic and/or
thermodynamic nonideality.  An essential part of this discussion is
selection of models.

For example, knowing you have a monomer dimer system often stems from
having a coiled-coil domain that must induce dimerization.  The problem
is you don't know 1) if it actually works in the context of your
construct, 2) you don't know if its also prone to aggregation, 3) you
don't know if it actually makes a trimer or tetramer instead or worse
yet a mixture of all of them.  Yes I agree sed equilibrium analysis has
fewer parameters and is thus preferable on many levels.  But sed
velocity as a function of concentration gives you a picture of the full
process including concentration dependence of zones and the presence of
Junk.  If you have appreciable Junk then sed equilibrium will be harder
to interpret & often the velocity pattern offers selective quantitative
analysis of regions of the pattern in a way that sed equil will not.
Many of my systems are not stable over the time of an SE experiment and
thus SV is preferred.

Thus my opinion is methods should be combined to establish a starting
point for analysis, a model, a hypothesis.  It is also my experience
that for well behaved systems (I wish I had more of them) you will
necessarily get the same answer, the same Kd or energetics, from direct
fitting of SV and/or SE data (the caveat being the extent of
hydrodynamic nonideality).

One additional point - Bo mentioned doing experiments above the Kd,
which of course requires knowing the answer in advance, or in fact doing
experiments over a wider and wider range of concentrations to achieve
that goal.  3 mm centerpieces have become common tools in my lab for
this reason.  The problem often encountered is that you don't go high
enough to populate additional species and you stop at dimer in the
fitting when in fact tetramer or octamer may also be there if you push.
It is instructive to make species plots where the weight fraction of
species is plotted vs the total concentration where you can see how well
you span the Kd.  

-------------------------------------------------------------------
Dr. John J. "Jack" Correia
Department of Biochemistry
University of Mississippi Medical Center
2500 North State Street
Jackson, MS  39216
(601) 984-1522     
fax (601) 984-1501 
email address: jcorreia at biochem.umsmed.edu
homepage location: http://biochemistry.umc.edu/correia.html
dept homepage location:    http://biochemistry.umc.edu/
-------------------------------------------------------------------



_______________________________________________
RASMB mailing list
RASMB at rasmb-email.bbri.org
http://rasmb-email.bbri.org/mailman/listinfo/rasmb


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.rasmb.org/pipermail/rasmb-rasmb.org/attachments/20040220/c7df2e65/attachment.htm>