TFA and basic peptides

[Mattias]

Peptides containing basic amino acids are usually analysed with TFA in the mobile phase to get the pH low enough to remove the silanol effect.

But if you want to separate the peptide from some of the deamidated forms (eg Glu, Asp) it is not so good to have low pH (then all species will all have the same charge).

Is it possible to use TFA at a higher pH (e.g. pH 4-7) and still get the "peak shape improving effect" of TFA? For instance adjust the pH of ammonium acetate with TFA to get pH 4?

[Andy Alpert]

Trifluoroacetate ion will form ion pairs with basic residues but not with acidic ones like Glu and Asp. Therefore, you're still probably going to get a deamidated peptide coeluting with the original Asn-containing peptide. If you want to separate the Asn, n-Asp, and isoAsp variants of a peptide, then use ERLIC. It separates them widely, with the n-Asp and isoAsp peaks in a characteristic 1:2 ratio. Ref: P. Hao et al., J. Proteome Res. 11 (2012) 1804. This works as well for peptides deamidated at Gln.

[Mattias]

There is no problem to separate the deamidated forms from the main peak at pH 5-6. Then the deamidated species elute many minutes before the main peak (because they are deprotonated).

But the peak shapes are terrible, especially for the peptide I am working on now. I have used a Waters BEH column and I get tailing factors of 5-6. Using TFA at pH 2 there is no tailing, but little separation.

The classical way would be to add some tetrabutylammonium sulfate, but that does not feel so "2014"

The ERLIC columns appear to be similar to the Primesep columns? That could be a way to go.

[DJ]

There is no problem to separate the deamidated forms from the main peak at pH 5-6. Then the deamidated species elute many minutes before the main peak (because they are deprotonated). But the peak shapes are terrible, especially for the peptide I am working on now. I have used a Waters BEH column and I get tailing factors of 5-6. Using TFA at pH 2 there is no tailing, but little separation.

The classical way would be to add some tetrabutylammonium sulfate, but that does not feel so "2014"

How so?

0.1% TFA lowers the pH, which helps suppress silanol ionization. TFA also ion-pairs with basic amino acid side chains (to increase retention). At pH 4-5, TFA will still ion-pair with basic side chains. A portion of silanols will be ionized at this pH, and the degree of ionization ("exchange capacity") will not remain constant over the course of the run.

The ERLIC columns appear to be similar to the Primesep columns?

No. Not at all. ERLIC was pioneered on a silica-based weak anion exchange column called "PolyWax." That is not to say P-sep columns could not work in "ERLIC mode," just that the stationary phase chemistry is dissimilar.

RP-HPLC works well to separate peptides that differ in hydrophobic AA content, but HILIC, ERLIC, IEX may be better options for separating Asp/Asn peptides.

[danko]

I use the anion exchange mode when I have to separate deamidated forms.
It works fine and is predictable.

Best Regards

[Victor]

"But the peak shapes are terrible, especially for the peptide I am working on now. I have used a Waters BEH column and I get tailing factors of 5-6. Using TFA at pH 2 there is no tailing, but little separation."


What is the mobile phase that you are using at pH 5-6 which is giving you such terrible results?

[Mattias]

Just ammonium acetate (10 mM) and acetonitrile. But there must be something about this peptide that makes it worse than others.