Predicted vs actual MW

Dima Klenchin klenchin at
Mon Nov 15 20:35:59 EST 1999

In article <80f0t5$5ae$1 at>, Kresten <kresten at> wrote:
:In article <pxpst2+-0611991043130001 at>,
:  pxpst2+ at (Peter) wrote:
:> In article <3821AEEE.9EFC3D58 at>, Lee Hunt
:> <L.hunt at> wrote:
:> Always remember that PAGE does not seperate purely by mass/size.
:> Seperation is by the charge to mass (m/z).  Alter this and and you
:> will have abherent migration.
:This is why we include SDS in our PAGEs. If it were so that proteins
:bound a number of SDS molecules proportinal to their mass and that this
:number was so large that the proteins own charge would be negligable
:compared to the charges that SDS bring, then all proteins in SDS would
:have approximately the same m/z. The idea is then, that you separate on
:m (or rather volume or some other structure parameter) using the sieving
:effect of the PA gel.

All is right but it's worth remembering that nothing is ideal under the moon.
In case of SDS PAGE, the number of SDS molecules bound per MW is
_not_ the same for all sequences. Thus, the mobility. The bottom line is
that SDS PAGE as a criterion for estimation of MW is better than 
gel filtration but can still be very far from true MW based on sequence - 
even when corrected by known posttranslational modifications. Some 
examples: phosphorylation (32+3x18 Da) at single site can upshift the 
band by as much as ca 5 kDa, calmodulin runs 18 or 22 KDa depending
on the presence of Ca++, myristoylation of ARF1 upshifts the band on 
a gel, but myristoylation of ARF6 (95% homology, but a very different pI) 
results in downshift. 


More information about the Proteins mailing list