Song Tan tan at mol.biol.ethz.ch
Sat Nov 25 09:08:57 EST 1995

In article <490bfg$j38 at nrcnet0.nrc.ca>, James.Fethiere at BRI.NRC.CA (James
Fethiere) wrote:

> Hello everybody,
> Are there any strong believers in Dynamic Light Scattering for estimation
> of the aggregation state of proteins?
> My problem is the following:
> I have a 18 kDa protein (on SDS-PAGE) that i concentrated to 
> 6.2 mg/ml.  This sample was diluted to 1.2 mg/ml in Tris-Hcl 20mM
> pH 7.4, 0.1M NaCl and ran on the dynamic light scattering machine
> after filtration through a 0.1 micron filter.  The results show an
> estimated MW in the range of millions and a polydispersed sample.  Once
> recovered it was filtered through a 0.02 micron filter;  the count rate
> dropped by about 25%, and the estimated MW was in the 20kDa range with the
> sample being monodispersed.  So, great!! my original sample is highly
> i suppose, and that is why i do not get crystals.
> But then, when i take the same sample at 6.2 mg/ml and run it on a SEC column
> (Ultraspherogel SEC 2000 from Beckman) in the same Tris+NaCl buffer the
sample elutes
> at a monomeric MW, either when injected at 6.2 mg/ml or when diluted at
1.2 mg/ml.
> No sign of aggregation whatsoever.
> So which experiment should i believe, and do any of you have any suggestions
> to confirm these results.

I guess I would believe the dynamic light scattering results.  Since high
molecular weight species scatters light much more than smaller molecules,
dynamic light scattering is very sensitive to larger aggregates.  So a
small blip at the void volume on an SEC column might show up as a sizable
signal by dynamic light scattering.

That having been said, I would have expected at least a small blip on the
SEC column if you had aggregation in your sample.  My general experience
has shown a very good correlation between dynamic light scattering results
and SEC for detection of aggregates in maromolecular samples.  And dynamic
light scattering is quicker to perform than SEC.

Is it possible that the sample contains some non-protein high molecular
weight species that is passing through the 0.1 micron filter, but not
through the 0.02 micron filter?   I guess I'm a little surprised that
simply passing the sample through the 0.02 micron filter was sufficient to
produce a monodisperse solution with the correct molecular weight.  This
suggests that your sample contains only two species:  very high MW
aggregates and monodisperse molecules, with nothing else inbetween.  If
this is true, then you might want to filter your protein solution through
a 0.02 micron filter just prior to setting up crystallization trials.

One possibility to explain the SEC result is the following:  the high MW
aggregates get trapped on the column (or guard column or prefilter) and
never elute.

Song Tan
Institute for Molecular Biology and Biophysics
ETH-Honggerberg (Swiss Federal Institute of Technology)
8093 Zurich, Switzerland
email:  tan at mol.bio.ethz.ch

More information about the Xtal-log mailing list