An interesting Peptide HPLC

[ananda]

Dear All,

I want to share this very intersting RP_HPLC chromatographic method we are validating at the moment. I woul like to hear comments and rationale for this chromatographic separation from experts here.

We have developed this RP-HPLC method to separate 2 peptides that differ by only 3 amino acids (at C-terminal). pIs of the 2 peptides are 4.6 and 4.7. We are using a tetrabutyl amonium IPR (1mM), and Mphase contain (B) 80%ACN and Mphase A's pH is 4.0 (just 1mM TBA in water) Just to remind you the pKa of alpha carboxyl groups of amino acids is also ~3.9,

If you think about all these pH values, it was surprising somewhat that we achived this separation but I could explain it in my validation report.. Can someone come up with a good rational for this separation at this pH.4?

Then at the end I will send my rational for the separation. Please dont make me re develop and validate this method..LOL

Cheers!
ANA

[HW Mueller]

Hm, another game, but very lopsided and rather useless unless you tell us what the column is, what the different amino acids are and how this difference relates to the retention time, how you know that your chromatogram is due to the peptides and not some impurity in them......

[Kostas Petritis]

Ananda,

Depending on the peptide amino acid composition, it's length and the nature of amino acid substitutes, the peptides might have been resolved just from the difference of their hydrophobic properties. You seem to imply that the whole mechanism is based on ion-pairing chromatography which might not be the case.

I have personally done some work on the peptide retention time prediction in RPLC and we have seen that the two amino acids in the edges of the peptide (N and C terminus) participates more in the peptide retention time than peptides in the middle (that might not be the case for very small peptides). For some general background on this see:
K. Petritis et al. Anal Chem. Anal. Chem. 75 (2003) 1039-1048 Use of artificial neural networks for the accurate prediction of peptide liquid chromatography elution time in proteome analyses.

As Hans said we will need more information to figure out any theories of why you see some selectivity between these two peptides.

[ananda]

Hans and Kostas,

The last 3 aminoc acids are Ile, Thr and Asp. Yes, Asp should play a role in the selectivity if there is any ion-pairing but at this pH, I also think there is no or minimum ion-pairing interaction for the selectivity.
The column is TosoHaas's Polymeric RP-HPLC column (2PW). The reson to use this polymierc column was very strong binding of TBA in silica columns reducing the life time.

I think there are 3 possible interactions that play roles on selectivity such as Ion-pairing, charge repulsion and hydrophobic interaction. So I think by "fine tuning" the pH and the ACN gradient we acheive the selectivity. Make sense?Please comment!

best Regards!

Ananda

[tom jupille]

How big (long) is the peptide?

[Chris Pohl]

ananda,

We are still missing two important pieces of information: the specific amino acids which are present in the place of the three amino acids you mentioned (Ile, Thr and Asp) and the elution order of the two peptides (i.e. which elutes first: the one with the pI 4.6 or the one with the pI 4.7).

[Kostas Petritis]

Ananda,

It would be useful to know how you calculated your peptides pI. If you did it by using web softwares be aware that most of them calculates the pI without any corrections.

As you might now the charge of a protein/peptide at a particular pH (when 0= pI) is handled by the charge determination factor. This is equivalent of the fractional charges of the protein's charged groups:

Z= Nterm + Cterm + a*K + b* R+ c*H + d*D+ e+E + f * C+ g * Y where Nterm, C term, K, R, H, D, E, C, Y are the charges these groups (amino acids) take on at a particular pH and the a,b,c,d,e,f,g are the count of each amino acid residue from the protein/peptide sequence. The algorithm is not perfect as it assumes that each group charge is independent of all others which is not quite right as if one positive charged amino acid (K,R,etc) is next to a negative charged amino acid (D,E etc) each probably does change the other's ability to take on a charge. Also if the peptide folds in a way that one positive charged amino acid is close to a negative charged amino acid then again each probably does change the other's ability to take on a charge.

So the real pI of your peptides might depend on their structure...

[ananda]

peptides are ~ 40 and 43 aminoacids. pI of 4.7 elutes first and then the other with pI of 4.6.

Kostas, thanks a lot for your comments but theoratical pI (4.6) and experiemental pI (by IEF, 4.2) are pretty close. I agree wioth your argument and these peptides dont really have a 3ry or 2ry structure but a mainly random coiled.

What I think is selectivity is due to hydrophobic interaction and charge repulsion rather than ion-pairing. Isnt it?

Thanks for all your comments!.

Ananda

[Kostas Petritis]

ananda,

the specific amino acids which are present in the place of the three amino acids you mentioned (Ile, Thr and Asp)

Ananda,

We still missing the above information...

[HW Mueller]

Ananda, we still don´t know how the varying amino acids are distributed between your two peptides.

[tom jupille]

Let's see, you can separate two peptides which differ by about 7% in length.

Unless I'm missing something, I don't see anything special here.

[Chris Pohl]

ananda,

Since you have the least cationic peptide eluting last, this would indeed tend to suggest that electrostatic repulsion was at least partially the basis for retention. But knowing how much was due to repulsion and how much was due to hydrophobic retention is difficult to estimate without knowing the differences between the amino acid residues of the two peptides.

One question, though: why not just separate the two analytes with a TFA containing mobile phase instead of operating under conditions where repulsive forces are playing a strong influence on retention? One would think this would give you more flexibility in controlling your separation.

[ananda]

Chris,

Thanks for your comments and I am agree with you that electrostatic repulsion and hydrophobicity are the main key selectivity factos rather than ion-pairing. We have tried TFA, and some other additives and IPR s as well and it came up that TBA was the only IPR that was able to selctively resove these 2 peptides even though the working ideal pH shows not much ion-pairing there.

Sorry for confusing others I thought my original message was clear when I said "C-TERMINAL. So the differnce between these 2 peptides is one has the last 3 amicno acids (Ile, Thr, Asp) the other doesnt.. The location is at the C-terminal and amicno acids are 41st, 42nd, 43rd aminoacids.

Tom, here we are trying to rationalize this selectivity especially at this pH of the method (4.) because pI of the ppetides and also pKa of alpha -COO- of acidic amino acid s are also ~4.0.

Thanks for your interest/comments!

Ananda

[Chris Pohl]

ananda,

Sorry for always more questions but which one is the longer peptide, the one with the pI 4.6?

[ananda]

Chris,

yes you are right. pI (theoratical) of 1-43= 4.6 and pI (theoratical) of 1-40=4.7.


Ananda