From: Geoff=Howlett%BioChem%UNIMELB@muwaye.unimelb.EDU.AU
  To  : tml@christa.unh.edu
  Date: Wed, 18 May 94 9:40:17 +1000

re: Re: Differential sedimentation.

Open letter to Tom Laue

My experience with differential sedimentation velocity is from a while back 
and is as follows:

There are basically two methods.

method one:  Use of wedge windows and schlieren optics.  This method accounts 
for the overwelming proportion of delta S experiments.  The main reason is 
that it is not as sensitive to small amounts of convection in the plateau 
region.  Aspartate transcarbamylase undergoes a conformational change on the 
binding of substrate analogues (about 3%) and this can be titrated with 
accuracies of 0.1%.  The secret here is to have good centerpieces (The ones 
from Beckman have improved out of sight in recent years (Personal 
communication , HK Schachman).  The other point is to align the cell in the 
rotor accurately.  My understanding is this is presently done on a trial and 
error basis where the line underneath the cell is calibrated with respect to 
the line on the rotor.  The ultimate test is to get a zero delta s for 
identical solutions.   Small corrections have to be made to the measured 
delta S for the effects of added ligands on density and viscosity.  These 
corrections are standard and not large, at least in the case of high affinity 
ligands.  Method one does not require absolute matching of menisci.

method 2:  This method was originally described for interference optics but 
has I believe been adapted for the XLA optical system.  It involves filling 
the reference sector with the protein sample and the sample sector with the 
protein plus ligand sample.  It is not necessary to exactly match menisci but 
they shold be relatively close together because the data appears as a peak 
due initially to the mismatch of menisci.  This peak then gets either bigger 
or smaller depending on the relative velocities of the sedimenting protein in 
the two sectors.  This method is much more sensitive to convection in the 
upper plateau but could well be preferred with the XLA.   

One of the interesting adaptions which needs to be made for method two in the 
XLA is to disable the auto gain capacity of the optical system.  I believe 
the XLA system adjusts the gain (voltage) on the photo multiplier due to 
changes in the OD in the reference sector at different radii.  You don't want 
this to happen in delta S experiments. 

How to disable this function?  I don't know.

As you correctly point out the delta S method is an extaordinarily sensitive 
method for picking up global conformational changes and is potentially a 
powerful tool to studies the effects of ligand binding and single amino acid 
changes on overall protein structure.

I'm not sure how we got here from sloping baselines.

Cheers

Geoff H.