Buffers in RP-HPLC of small peptides

[annaand]

I'm trying to estimate adsorption isotherm of tripeptide (Leu-Leu-Leu) using frontal analysis (FA) and FA by characteristic point (FACP) methods. Column: Jupiter Proteo (4.6x150mm, 90A, C12, end-capped), mobile phase: 48%:52%:0.1% MeOH:H2O:TFA (v/v/v) I came across the problem with keeping constant pH for the whole range of peptide concentration. Basically, I have 3 solutions of LLL (in mobile phase). I measure breakthrough curves of each solution mixing 5%-100% of the solution with mobile phase. I get 11 breakthrough curves. I do the same for all 3 LLL solutions. At the beginning, when the concentration of LLL is small, the amount of TFA is sufficient to keep constant pH. But when I increase the concentration of the peptide I get breakthrough curves with shoulders and pH of this LLL solution is higher than the previous ones. That, I guess, means that there is no enough of TFA to keep one form of LLL. Since I want to measure adsorption isotherm I have to know what I am adsorbing or at least be sure that it is the same form for the whole process.
Now there are my questions: Should I use other buffer than TFA? What buffer should I use? Is it reasonable to change concentration of TFA going from one LLL solution to another?

[Mark Tracy]

Try converting the LLL to the trifluoroacetate salt before making up the samples. Titrating your stock solution of LLL might be the easiest way.

[Uwe Neue]

Yes, you hit the problem on the head...

The large amount of peptide changes your environment. A conversion of the peptide to the TFA salt may help, but it is not a guarantee either, since nothing forces the TFA to hang around, as your LLL adsorbs on the surface. In addition, you are effectively changing the TFA concentration in your isotherm, which is a function of the method that you are using. In my opinion, you should check, if the profiles observed or the measured isotherm parameters are changing as you increase or decrease the TFA concentration in the background mobile phase (~5x change).

If you want, we can chat more about it in private: Uwe.Neue@prodigy.net

[HW Mueller]

Can´t you do the whole experiment with concentrations of LLL which are in the range that does not effect the pH? I don´t see how you can compare isotherms determined in different media.

[annaand]

Thank you very much for your replies. They are very helpfull.

Mark: Titration to obtain LLL TFA salt sound reasonably. Let's assume that did it. Then I know the amount of TFA added to sample ("concentration" of TFA in my LLL solution). What about background mobile phase? Should the conc. of TFA be the same there as in the sample? I guess so. This way I will change the TFA conc. going from one LLL solution to another. And that's what I wanted to do but I am still not sure if it's reasonable.
HW Mueller: I already did what you suggest. But I am still in linear part of isotherm. And this does not tell me much about the mechanism of adsorption process.
I have one more question which came across my mind. What about the isoelectric point of LLL. It is 6.01. Is it thoughtful to use phosphate buffer to keep pH=6? Any predictions?
Thanks very much for help!
Anna

[Mark Tracy]

Yes, you should adjust the total TFA concentration in the sample to be the same as the mobile phase.

A phosphate buffer for the sample and TFA in the mobile phase would make the problem worse. If you want to use phosphate in both sample and mobile phase, that is different.

[annaand]

Exactly I wondered if phosphate buffer would help (meaning in both stock solution of LLL and mobile phase). But what I can expect is probably the same problem. When the conc. of LLL is big, there won't be enough of buffer to keep pH. Unless I will put "a lot" at the very beginning (meaning for small LLL conc. too).
Of course you can say that I can check everything (which probably in the end I will have to do) but the truth is that LLL is not so cheap and I don't want to waste much amount of this peptide. That's why I'm trying to find theoretically the best way to finish this step.

[Mark Tracy]

The retention mode for phosphate buffers will be almost pure reversed-phase (assuming you have a really good C18 column). With TFA mobile phases, there is a small but significant contribution from ion-pairing effects. Phosphate is practically insoluble in the C18 phase, whereas TFA is somewhat soluble.

[annaand]

That's a really good point! I haven't thought this way. I was also considering to use organic buffer (triethylamine/acetic acid, pH~6) But you've made a really good point. Of course I want to have one adsorption mechanism as possible. I'd rather not to have additional ion-pairing effect. Though it is very interesting how it will change the mechanism of process by changing e.g. a character of buffer (inorganic, organic) with the same pH.
Thank you very much! You are really very helpful.
Anna

[HW Mueller]

I am not so familiar with isotherm determination, so I don´t see why you want to get into nonlinear regions. In your initial statement you mention that you want to have one species of LLL. Won´t you have differnt entities (aggregation etc.) when getting into nonlinearity? Even if you keep the pH constant?

[annaand]

Yes, I want to go further than linear region of isotherm. The best would be to get plateau. Why? Because this way I can predict the mechanism of adsorption using some adsorption isotherm equation to fit experimental data and on this basis conclude about the mechanism of adsorption LLL onto hydrophobic surface. E.g. Langmuir equation - ideal (without lateral interactions) adsorption onto homogeneous surface, Freundlich - ideal adsorption onto heterogeneous surface, Fowler - nonideal onto homogeneous surf, or some other.
That's what I want to do - to measure adsorption isotherms of small peptides onto C18 (or C12) by means of many methods (FA, FACP, ECP, IM, PM) and compare the efficiency and accuracy of them. Of course all of this method, if their results will be compared, should be made in the same conditions: mobile phase composition, column, temperature, pH etc.
Anna

[Mark Tracy]

To make sense of your work, you will need a well-characterized reversed-phase column. It should have to lowest possible silanol activity to minimize ion exchange interactions; a new column never exposed to low-pH conditions. The USP database might be a good place to compare columns http://www.usp.org/USPNF/columns.html. In fact, you probably will be able to use your technique to characterize stationary phase surfaces of different columns.

[annaand]

Mark,
I have brand new column from Phenomenex-Jupiter Proteo. Phenomenex says that this column is:
1. for HPLC, LC/MS, Prep-LC, Capillary-LC, 2-Dimentional Chromatography of Peptide maps and Peptides (< 10,000 Daltons),
2. Phase: C12 (90Ã…) with proprietary, non-polar endcapping
Unfortunately, I couldn't find it listed in your link. But as phenomenex says it is endcapped what means no polar (silanol) groups presented on the surface, right?

[Mark Tracy]

I have not used that column, so I can't comment on it. Call Phenomenex.

[Uwe Neue]

If you make a TFA salt, you will need to consider if the TFA will migrate with the LLL or not, and how this will affect the profiles and the isotherms that you are trying to measure.