Hi, Tassos,
It appears I am no better a biophysicist than you are on this one (I can
see my Scattering instructor from grad school disapproving of me right
now, he reads this bb). I cannot connect how the laser-induced dipole
and concentration-dependence of refractive index of your sample would
depend on each other (any biophysicists out there?).
But I still thought I should mention a few practical points that might
help: dn/dc depends a bit on wavelength, so not all literature values
might be the right ones for you (we use 0.185, because someone at Wyatt
must have told us so). Second, if you really want to know your dn/dc,
measuring it might be the thing to do. You probably already have an RI
detector (Optilab from Wyatt?). If you can get yourself a medical-grade
syringe injector, you can inject your sample into the RI detector
(otherwise it is very hard to do so) at multiple concentrations, and
calculate for yourself a dn/dc. Obviously, you have to (1) have decent
quantities of material to waste, and (2) be very accurate with your
dilutions/concentrations. Wyatt also sells something that might help you
do this experiment, that might be the easier way.
Third, sometimes, inspection of the Debye plot and removing outlier
several angles from the plot helps improve accuracy (and precision, but
that is not always a good thing). Especially smaller angle measurements,
which seem to sometimes cause trouble in our hands (We usually use
detectors 10-17 on the 18-angle Wyatt MALS machine). This sounds like
throwing away data, but it is more like throwing away outliers and
inaccurate data (Wyatt claims only three data points is enough to get a
good molar mass, in most practical cases). Use your statistical judgment:)
Good luck,
Engin
On 3/30/09 1:16 AM, Anastassis Perrakis wrote:
Hi Engin -
On Mar 27, 2009, at 15:57, Engin Ozkan wrote:
Dear Tassos,
Your assumptions are right, if (1) your dn/dc is accurate, or (2) your
machine is calibrated.
indeed!
We recently measured a protein of a similar size
to yours, and when a 700 Da ligand was added to the buffer, the measured
protein mass was increased accordingly. So MALS can be pretty accurate.
For our dn/dc, for pure proteins, we always use 0.185 (not 0.19).
yes, thats what most people use. Some literature claims 0.182 ... 0.19
is an old value, indeed.
For
sugar groups, we assume a dn/dc of 0.14, and estimate a mass-averaged
value for the glycoprotein (usually somewhere between 0.175 to 0.18).
For DNA and RNA, the values will be different, again.
yes.
You may also realize that by changing a simple calibration constant, you
can modify your measured molar masses anyway you want. It may be time
for a recalibration (it is not difficult, you can do it yourself). We
tend to regularly run BSA, and see if everything is as expected with our
equipment.
yes, we did re-calibrate. I was very puzzled since we had very
accurate results before as well.
my question was more towards trying to understand my protein a bit:
the fact that it appears to have a strange dn/dc value, would it
suggest that it is a
very strong molecular dipole?
I better solve the structure ;-)
Cheers - Tassos
Good luck,
Engin
Anastassis Perrakis wrote:
Dear all,
The MALLS instruments on-line with an FPLC and with an RI detector,
should provide an 'absolute MW', shape independent,
and indeed in our hands they do well. Until yesterday, where a 21kD
protein pretends to be 25 kD. We did the mass spec
anyway, and its 21kD as we expected to the residue, but I am still
puzzled by that result.
One central assumption for the MALLS formulas, is that dn/dc, the
specific refractive index increment, is constant for unmodified
proteins,
made by aa with no sugars etc. Literature suggests dn/dc values for
proteins to be constant and between 0.189/0.190 is a good value,
with minimal buffer dependence for aqueous buffers with 'the usual'
salts.
I am a rather bad physicist, but my reading tells me that dn/dc, and
thus light scattering, depends to the "laser-light induced dipole in
the molecule". Is there any reason to believe that in theory a
molecule with a very particular charge distribution (eg a small DNA
binding protein which is already a 'dipole') would have significantly
different dn/dc values? Is anyone aware of such an experiment?
Literature searches were in vain ...
Best -
Tassos
*P** **please don't print this e-mail unless you really need to*
Anastassis (Tassos) Perrakis, Principal Investigator / Staff Member
Department of Biochemistry (B8)
Netherlands Cancer Institute,
Dept. B8, 1066 CX Amsterdam, The Netherlands
Tel: +31 20 512 1951 Fax: +31 20 512 1954 Mobile / SMS: +31 6 28 597791
