From: Jim.Bloom.B@bayer.com
  To  : -         (052)rasmb(a)bbri.harvard.edu <rasmb@bbri.harvard.edu>
  Date: Mon, 17 Feb 1997 11:41:16 -0500

Difference sedimentation

I have received a number of replies to my question about difference
sedimentation (see below) and in one reply from Walter Stafford he asked that I
re-transmit them to the group as he felt that a number of people would be
interested.  As I have cleverly deleted them but fortunately printed them out I
will give a summary of the responses.  Marc Lewis, Claus Urbanke and Peter
Lavrenko all suggested making fine communicating channels between the two
sectors at the top and the bottom (Claus suggested taking some sort of needle
and making a small scratch at the very bottom and the very top of the
centerpiece connecting the two sectors.  The scratch should be rather fine,
such when filling the cell the solution cannot creep from one sector to the
other.  When running however the menisci will equalize and there will be a
small amount of solution flowing at the bottom from one sector to the other)
small mixing at the bottom should not be a problem or as Marc said put a few
more microliters of protein without DTT in one sector and thus a very small
amount of non-DTT protein flowing across at the bottom to equalize the menisci
but also will avoid getting fairly diffusable DTT into the other sector.  Marc
also suggested putting a few microliters of FC-43 in each sector to give the
same result but also further ensure no communication between the two protein
samples.  As I am not quite as fearless as Claus with a needle and I am trying
to contact someone at Beckman about making the channels.
 Other responses came from Helmut Coelfen who determines particle size
distributions in an 8 hole rotor with turbidity optics where the top of the
cell defines the meniscus.  He fills the sectors completely but finds it tricky
to fill the cells without any air bubbles and suggests shaking the cells while
continuously injecting sample so that the liquid (and sooner or later the air
bubbles as well) leave the filling hole.  He has not done difference
sedimentation this way but suggests trying it.  Mike Jacobsen has another
approach to difference sedimentation described in Biochemistry 1996, 35,
13173-13179 (Jacobsen, Wills and Winzor) and he faxed a preprint of another
paper  by Jacobsen and Winzor "Studies of ligand-mediated conformational
changes in enzymes by difference sedimentation velocity in the Optima XL-A
ultracentrifuge".  The Email address on the preprint is
winzor@biosci.uq.edu.au.  Finally Walter Stafford wrote to say that he has
worked out a method to get delta-s from a difference experiment using the g(s*)
software.  He uses a synthetic boundary cell to match the meniscuses.  The
extraction of delta-s requires some curve fitting...he gave me his phone number
for details (I am guessing it is ok to send it out, forgive me Walter if I've
goofed here!!) 617-742-2010 x 386.
  I would like to thank all who responded and RASMB-how did we survive before
the Internet??

Jim
---------------------- Forwarded by Jim Bloom/BRKL/PH/US/BAYER on 02/17/97
08:10 AM ---------------------------


Jim Bloom
02/12/97 08:57 AM
To: rasmb @ bbri.harvard.edu @ internet
cc:
Subject: Difference sedimentation

I would very much appreciate some advice!!  I have been interested in the
effect of DTT on the structure of my protein and have determined the
sedimentation coefficient of the protein with and without various amounts of
DTT and observed a small decrease in S which makes sense to me.  I did many
repeat runs but the difference in S was so small that the results weren't
terribly convincing so I tried the Schachman difference technique (with the
E-mail help of Tom Laue-thanks Tom!!!) by which I mean protein in one sector
and protein plus DTT in the other sector of the same cell with the same volume
in each sector.  I got very nice results with this technique and am generally
happy except for one thing...I found it very very difficult to get the two
menisci exactly matched!  some times they would be ok and sometimes slightly
off and the data would be no good.  I tried several pipetting techniques and am
convinced that I am within one microliter out of 400, yet eyeballing the
menisci's seemed to show they were not identical yet in the XLA at speed
(60,000) sometimes they would be nearly identical and sometimes I would see a
gaussian right off the bat...so I tried filling exactly and then adjusting by
eyeball and then things were worse.  Maybe all of this is confusing but what's
up?? are the sectors somehow distorted or not exactly the same??  Does one
sector get distorted at speed diferently than the other or am I just all
thumbs??  I really love this technique both in theory and in practice but when
one in three or two in three of your runs aren't any good one gets tired and
gives up!!  Any advice?

Jim Bloom
Bayer Corp.
Berkeley
510-705-5301
Fax:  510-705-4940