using UV absorbance scans to 'measure' homogeneity of proteins

Peter Prevelige prevelig at
Fri Sep 10 07:27:31 EST 1999

Achim Recktenwald wrote:

> If you are looking for non-covalent aggregates, i.e., no intermolecular
> bonds like SS-bridges, you can in many cases check  your protein by simple
> UV280nm measurement.
> Make several dilutions of your protein over a wide range, e.g., 5fold up to
> 500 or 1000fold,  then measure your absorbance. If your total absorbance
> (absorbance x dil.factor) does not stay (more or less) constant for all
> dilutions but increases,  you have a good indication for protein-aggregates
> in your original solution - assuming you made sure their is no other
> particular matter in your protein or buffer solution.
> Often you can observe that the total absorbance increases up to a certain
> dilution factor and then stays constant.
> The aggregates tend to break up upon higher dilution, and you 'find' more
> protein.

this strikes me as curious. i wouldn't expect aggregation to affect the
absorbance spectrum per se. i will grant that there could be absorbance
changes upon association, in much the same manner that there are
absorbance changes upon folding/unfolding. these types of changes which
form the basis of UV difference spectroscopy are generally fairly small. 

furthermore, dilution will dissociate aggregates, which will decrease
scattering, particularly at 280 nm (given the relatively steep
wavelength dependence of scattering) so it seems likely that dilution
may actually disproportionatly decrease rather than increase the
absorbance if there are aggregates, at least in the general case. 

does this method really work? 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: prevelig.vcf
Type: text/x-vcard
Size: 385 bytes
Desc: Card for Peter Prevelige
Url :

More information about the Methods mailing list