From: Allen P. Minton <minton@helix.nih.gov>
  To  : rasmb@bbri.eri.harvard.edu
  Date: Wed, 22 Jun 1994 16:02:53 +0059 (EDT)

simultaneous determination of M and vbar

My colleague Saleh Darawshe has been attempting to measure M and vbar of
an unknown protein via the technique of measuring the buoyant M in
solvents of differing density.  He prepared buffers in 100%H20, 75%H20 +
25%D20, ......, 100%D2O, making 5 in all, and measured the density of each
on an Anton Paar densitometer.  Solutions of the unknown protein were
dialyzed against the buffers and subsequently centrifuged to sedimentation
equilibrium.  The experimental gradients were well fit by an ideal sed.
equil. expression, yielding a best-fit value of M*, the buoyant density,
for each buffer.  The plot of M* = M(1-vbar*rho) was plotted against rho
and a nice straight line was obtained as expected.  The y-intercept is M
and the slope is -M*vbar.  The first indication that something was not
quite right was that the best-fit value of vbar was 0.62 ml/g.  This is
too low for almost any protein.  The investigators were able to provide a
sequence obtained from cloning of the gene, and from the AA composition
Saleh computed an approximate vbar of 0.73.  So I suggested that he do a
control experiment on ovalbumin, for which we know both M and vbar.  He
prepared ovalbumin in a similar set of buffers and centrifuged to sed. 
eq.  Once again, the plot of M* against rho yielded a nice straight line
(very well defined by the data).  However, this time the best fit
parameters were M = 38 K and vbar = 0.70 ml/g.  The 95% confidence upper
bound for each of these values was 41 K and 0.715 ml/g, vis-a-vis
literature values of 45 K and 0.745 ml/g.  Does anyone have an idea why
this technique is not working for us? 




Message-Id: <9406221631.AA05305@amgen.com>
Date: 22 Jun 1994 09:32:44 U
From: "JOHN PHILO" <JOHN_PHILO@amgengate.amgen.com>
Subject: Re: Aluminum center pieces
To: "Borries Demeler/Biophysics" <demeler>
Status: OR

        Reply to:   RE>Aluminum center pieces
Hi, Borries
     I originally purchased only aluminum centerpieces since the size of our
molecules almost always means I run at 60K.  Actually, I believe that they
may be anodized, but not in the presence of the black dye commonly used with
aluminum anodization.  (Since thin layers of aluminum oxide are transparent,
it is not obvious when this has been done.)  To me, the surface felt as
though it may have been anodized.  Anyway, we used them with phosphate
buffered saline and for a few experiments at pH 4 and nearly zero ionic
strength.  There did seem to be some change in the surface over time, but I
suspect it was due more to the detergents I used to clean them (which are
invariably quite alkaline) than the samples/buffers run in them.  Generally
aluminum is pretty inert at neutral or acid pH, but is attacked by strong
bases.
       I have now pretty much stopped using them.  For certain proteins, what
I found was that they were causing precipitation on the surface and the
appearance of aggregates in the solution.  Sometime I got some really bizarre
effects.  For example, with certain proteins when I would start the run and
take a scan at 3K I would see a dip in concentration near the middle of the
cell by as much as 20%.  If I would then quickly stop the rotor, remix the
contents, and start again, I would see a uniform concentration, but at a
concentration lower than what I expected.  Apparently what was going on was
that protein was precipitating on the surface along the regions on the face
of the centerpiece between the channels and the red gasket (there is a thin
D-shaped region of solution at each face due to way the gaskets seal).  As
this precipitation occurred, material from the channels was diffusing out
into these regions, depleting the concentration in the channel, with a
minimum concentration where the D-shaped region is fattest.  This effect
seemed to get worse as the centerpieces got older.
       Last December I borrowed some charcoal epon centerpieces and ran the
same material in the aluminum centerpieces and in epon.  I found that the
aluminum gave variable amounts of aggregates while the samples in epon were
aggregate free.  I immediately ordered epon ones and have not used my
aluminum ones since!  I know you are not supposed to take the epon to 60K,
but I do it anyway.

Regards, 
John