Hi, I'm doing Pepsin and Trypsin digests on a protein with a haem attached (covalently). The Peptides resolve very well, however, the haem appears as a very poorly resolved peak (typically 30 fold less resolved than peptides, even greater with trypsin digest).
We think it is because the haem of the haem-peptide is ligating or associating with the peptide chain of another haem-peptide, thus creating a dynamic system (or multiple strand/ring lengths).
Any comments on this?
Any suggestions as how to achieve good resolution?

HPLC conditions are:
C18 (4.6 x 250mm column) Gradient: 5%B rising to 60%B (1ml/min)
(A=0.1% TFA, B= Acetonitrile, 0.1% TFA)

Going with your hypothesis of association with another peptide, you might try a higher temperature (which should favor dissociation).

We make software (DryLab) that can model gradient shape + temperature effects based on 4 experimental runs, so we've had a chance to look at quite a few problem separations. You'd be surprised at the difference temperature can make (then again, maybe you wouldn't be surprised!).

If you'd like, I can take a look at your separation. I'd need the results (report + chromatogram) from 4 runs (say, 30- and 90-minute gradients at 35 and 50 degrees). You can e-mail me off-line if you're interested.

Thanks for the tip Tom.
I have tried several temperatures ranging from 20-60oC. There is virtually no effect on the peak width. Dispite great difference in the peptide elution profile (as expected) almost no difference was seen in the elution time of the haem 'smudge'. Could this mean something?
I have also tried adding SDS (0.05% - 0.5%) to the eluent to try to disrupt the association (if that is what it really is). This also had little effect.

Seperation is no problem, I can observe the haem at another wavelength to the peptides.

One possible solution may be to block the haem iron. Unfortuanly binding ligands like CO, isoCH3NC or N3 are no good as thay drop off in the HPLC.

Another possibility is that the Haems groups are staking, however, I don't see this with any other free haem.

Help

Interesting! Since you can do the quantitation at a different wavelength, is the problem a practical one (e.g., detection limit?) or an aesthetic one (the haeme peak is ugly).

In my experience, peaks that are anomalously broad are often the result of "microheterogeneity" in the analytes. In effect, you have a number of slightly different forms which are slightly separated. You have seem to have eliminated one set of possibilities (different conformers) with the temperature experiments. Is there a possibility that you might have a variety of bound species (different binding sites, different orientations, etc.?)? If so, the remedy is to try to force the product to a unique form.

I'm afraid I can't offer any more cogent suggestions than that. maybe someone else can chime in.

-- Tom Jupille