The pI is a theoretical value and may not reflect the true pI accurately. Nevertheless, by raising the pH you increase the negative charge on the protein (on average) so that it will bind with higher affinity to an anion exchange column.

Probably, you need to decrease the salt concentration in your Loading Buffer. Also, you need to make sure that your protein itself does not have any salt. The presence of salt in any of the cases will compete with your protein for the binding to DEAE resin.

There is no salt in the loading buffer sir. 

Protein samples are run through desalting column to remove any salts too.

What i dont understand is why is there a difference when the pH is 7.2 and when the pH is 8?

It is probably due to the ionic strength of your buffer. If it's too high, proteins won't bind to the ion exchange resin because the anions in your buffer will compete for the binding sites.

Try reducing the ionic strength or try changing the buffer altogether.

in that case the protein should not bind at pH 8  also?

The pI is a theoretical value and may not reflect the true pI accurately. Nevertheless, by raising the pH you increase the negative charge on the protein (on average) so that it will bind with higher affinity to an anion exchange column.

OK. Thank you sir. 

Hi there,

As Adam mentioned pI is calculated from protein sequence using theoritical pK for each protonable function. In the native protein, it might be different because of pK modification due to proximity effect between chemical functions. Furthermore the binding of a native protein to IEC column also depends on charges present at the protein surface. In your case, clearly the apparent pI of your protein is certainly higher than the theoritical one.

is there a way we can determine the experimental pI?

Yes : Isoelectric focalisation gives the best idea of actual pI. For info:

http://en.wikipedia.org/wiki/Isoelectric_focusing

Hi,

One interesting observation is that some titratable residues (eg HIS, CYS) will have pK values in the vicinity of  7-8. The exact values will depend on the local electrostatic environment. Therefore some of these residues will be changing their charge in exactly this pH interval. This may contribute to their affinity towards the separation technology you use.

In my understanding the distribution/density of charged residues on the surface is the primary determinant of binding to ion exchange resins and not just the PI (theoretical or otherwise). Certain proteins bind both anion and cation exchange columns robustly at the same pH. I agree with Adam and will go further and assert that the computed pI is many times simply a distraction when strategies are planned for protein purification. My preferred method is anion exchange first  at pH >8.0 (most proteins will bind). Then try cation exchange if necessary. Ignore the pI. If your protein is soluble and functional it can be purified some way.

I agree with the above comments. In addition I would like to mention that the plot of the charges vs pH of a protein shows very often a flat zone around the isoelectric point  where the number of charges does not change very much. This "flat zone" can extend over 1 or 2 pH units and this may be responsible for your observation. If you know the sequence of your protein you can use http://protcalc.sourceforge.net/ to test this hypothesis.

http://protcalc.sourceforge.net/

In response to Jean Michel Bruneau's answer :

Knowing the sequence may give you a first idea about titration behavior - but it is likely to differ from a more correct calculation, where the 3D structure is included. In the latter case local electrostatic fields will be included, and pending size of this contribution, the pK values will change. See e.g.

How do lipases and esterases work: the electrostatic contribution

By: Petersen, MTN; Fojan, P; Petersen, SB

JOURNAL OF BIOTECHNOLOGY Volume: 85 Issue: 2 Pages: 115-147 Published: FEB 13 2001

Protein engineering the surface of enzymes

By: Petersen, SB; Jonson, PH; Fojan, P; et al.

Conference: 8th European Congress on Biotechnology Location: BUDAPEST, HUNGARY Date: AUG, 1998

JOURNAL OF BIOTECHNOLOGY Volume: 66 Issue: 1 Pages: 11-26 Published: NOV 18 1998

Article: Protein electrostatics: a review of the equations and methods used to model electrostatic equations in biomolecules--applications in biotechnology.

Maria Teresa Neves-Petersen, Steffen B Petersen

[Show abstract]

Biotechnology annual review 02/2003; 9:315-95. DOI:10.1016/S1387-2656(03)09010-0

Have you checked the actual pH of the eluted fractions?. Is your pHmeter well calibrated? This should be  confirmed before looking for a reason. There are differences in between the theoretical and experimental pH, the interaction in between the protein and the ionic groups not only depends on the pH but also on the ionic strenght and other physico-chemical factors, there are other interactions.

I do agree with all the above answers.