From: SCHOENFH <HANS-J.SCHOENFELD@Roche.COM>
  To  : parker <parker@topaz.syntex.com>
  Date: Wed, 09 Jul 1997 16:08:04 +0100 (MET)

RE: Required purity for AUC

--> Hi! I would like to know people's views on what the minimum acceptable
--> purity level is for studying proteins by analytical ultracentifugation
--> (particularly sedimentation equilibrium). I try to obtain >99% whenever
--> possible; is a slightly lower purity likely to cause much trouble? (I
--> realize that it depends on the relative extinction coefficients of the
--> sample and contaminant.)
--> 
-->     Thanks!
--> 
-->     Matthew Parker


Dear Dr. Parker:

Your question can not be answered in terms of a given percentage. As you know 
people have different interpretations of the meaning of "the purity of a 
protein". Often purity is judged based on Coomassie stained SDS-PA gels. I 
suggest that the percentage of protein contaminants that are visible on such 
gels should be below the measuring error of the AUC technique. On this respect 
I suppose that you are on the safe side with your >/=99% limit. The situation 
might be different if a contaminant interacts in a catalytic manner with the 
protein under investigation.

However, contaminants of low molecular masses or nonprotein substances such as 
substrates or cofactors that are not visible in SDS-PAGE may influence protein 
quaternary structure even when they are present only in the submillimolar 
concentration range. We recently described such an example where minor 
quantities of ADP or ATP strongly influence the equilibrium of GroEL/S 
chaperone complexes [Behlke, Ristau and Schoenfeld (1997), Biochemistry, 36, 
pp. 5149-5156].

Another, perhaps more frequently observed, situation is contamination by 
misfolded protein molecules, which also may not be detectable by SDS-PAGE. 
This means that the purity of a protein sample has not only to be analyzed 
with respect to its sequence homogeneity (primary structure) but also with 
respect to its homogeneity in quaternary structure. During recombinant 
overexpression misfolded protein often aggregates or proteins containing 
cysteine residues often form artificial intra- or intermolecular disulfide 
bridges. We faced such a situation during our AUC investigation of DnaJ 
[Schoenfeld, Schmidt and Zulauf (1995), Progr. Colloid Polym. Sci., 99, pp. 
7-10].
Unlike specific oligomerization, unspecific aggregation is irreversible and 
time dependent. We investigate our protein preparations by measuring diffusion 
coefficients D using simple quasielastic light scattering (QLS) 
instrumentation such as the DynaPro 801 (Protein Solutions Inc.). If diffusion 
coefficients are broadly distributed or far smaller than expected for 
reasonable oligomers of the protein under investigation we evaluate the time 
dependency of D. If D increases significantly over a few days the protein 
obviously has an aggregation or a folding problem. QLS measurements are faster 
and simpler than AUC experiments and therefore can save a lot of effort and 
time if used as a quality control. Furthermore, the obtained diffusion 
coefficients are accurate enough to be used to determine molecular masses from 
sedimentation runs via the Svedberg relationship. If an aggregation problem is 
encountered we invest more effort into protein expression and purification 
before starting with AUC experiments.

With best regards,

Hans-Joachim Schoenfeld.


========================
Dr. Hans-Joachim Schoenfeld
c/o Hoffmann-La Roche Inc.
PRPI, B93/5.44
CH-4070 Basel
Switzerland

Tel. (CH)-61-68-82895
FAX  (CH)-61-68-89060
email: hans-j.schoenfeld@roche.com