larson eric elarson at
Tue Sep 12 12:50:06 EST 1995

sg at (Stephen Gately) writes:

>We wish to purify an enzyme using ion exchange from serum-free conditioned
>medium.  The NaCl concentration is approximately 100 mM.  We would like to
>desalt this medium prior to binding to DEAE-sephadex.  We have tried to
>desalt using buffer exchange (Centri-Prep), however, the enzyme loses its
>activity after desalting by this approach, even if NaCl concentrations are
>raised back to physiological levels!
You can desalt by dialysis, desalting gel filtration, protein ppt., or remove 
the ions by Sterogene Ionclear Bigbead resin.
Dialysis can be combined with protein concentration by putting polyethylene
glycol 10,000 (sigma) in the dialysis buffer (not in the bag) and by using
a dialysis bag that will exclude the PEG 10K (standard 10,000 to 12,000
cut-off works fine).
Gel Filtration requires a properly-sized column (generally 5 to 10X the
volume being desalted).  Sephadex G25-300 (approx 2,500 mol. wt. exclusion
with 300 micrometer beads) is generally used (but there are many others
Occasionally, precipitation of the protein can be used to advantage.  I
routinely ppt. proteins using PEG 10K (with 10 mM Mg++), to "reprecipitate"
proteins that were already ppt. by ammonium sulfate.  Ammonium sulfate ppts.
tend to have a large fraction of membranes for the initial cut (35% for our
plant protein).  However, the resuspended AS pellets often contain enough
ammonium sulfate to cause binding problems on ion exchange resins.  We've
found that many (not all), of our proteins will reprecipitate out of the
now dilute AS solution using 14 to 17% PEG 10,000 (same as used above
for dialysis/concentration).  The advantage is the supernatent of the PEG
ppt. contains the dilute AS.  Another distinct advantage is once the PEG
pellet has been resolubilized, it is found that the membranes are now *not*
soluble and can be spun out (I believe dilute AS helps keep the membrane
fragments in solution).  Essentially, doing a PEG ppt. on the heels of an AS
ppt. removes both the AS and contaminating membranes (but maybe in your
case the membranes are the wanted fragment?, fine, they could spin out now
with far fewer contaminating soluble proteins).

The Ionclear Bigbeads is a new technique I'm itching to try.  Product
literature claims that ions can be clear from solution without significant
protein loss.  I tend to believe product literature claims after I've tried
the technqiue (1-800-535-2284).

>Are there any other techniques available that may be used to desalt the
>conditioned medium?

It is often the case that simply diluting the protein sample containing
monovalents allows for protein to bind to the column in question.  DEAE is
a relatively weak binder, with some proteins eluting as low as 30 mM NaCl.
This would probably be a lower limit for diluting.  By altering the binding
chemistry to a slightly tigher binding group, the need for diluting can be
reduced.  I suggest trying QAE Sepharose (fast flow is usually best when
the sample is possibly contaminated or if this is the first column in the
series).  Sizing of the column can be a bit tricky, but for first trys, I
usually load about 3 to 5 mg of protein per mL of resin.  If the sample
contains significant divalents (sulfate from AS the principal problem),
then some significant increase in col. size is warranted.  Divalent anions tend
to stick to DEAE and QAE resins very strongly, to the point they almost
need to be considered as "permanent" binders for the loading phase.  This
can mean uneasily large columns for relatively small amounts of protein.

I have had occasion to use a bulk-column method for catching a rare protein
from dilute solutions.  Usually, I'll dilute the sample to monovalents to
around 50 mM (i.e. 50 mM NaCl), then add washed QAE resin and mix.  The
weak slurry is spun in a prepartive centrifuge (equal volumns in opposing
tubes as when the resin is spun, the balance of the tube is markedly
altered), the supernatent is aspirated, and the resin is resuspended with
minimal buffer and poured into a column that contains 25% of QAE already
settled (with excess buffer already aspirated).  This "column" is then
subjected to a rapid gradient (usually NaCl increases).
The goal with any protein that shows lability is to move quickly to get it
onto a column and away from possible degradative agents (i.e. proteases).
If it were me?  I'd dilute the sample to 50 mM NaCl, load it onto an
oversized QAE column, and do a quick gradient with deliberate speed at all
stages (especially the analysis phase, don't let the samples in the
fraction collector "just sit" as you do assays).

Good luck.

Eric Larson                  | University of Illinois at Urbana-Champaign
USDA/Agronomy                | 190 ERML; 1201 W. Gregory; Urbana, IL 61801
elarson at     | Voice 217.244.3079  Fax 217.244.4419
Fidonet: 1:233/4.1           | My opinions are my own, but correct :-) 

More information about the Proteins mailing list