Biotech Values

[poorgradstudent]

1) What is the chance that one peak in a CE graph ends up a part of multiple different peaks in an HPLC curve? And, if it can map to multiple peaks, how hard would it be to do the mapping?

For these types of analyses, a chemical may come out as multiple peaks in a CE or HPLC run if, in the analyzed mixture, it is present as a free form as well as a form that is bound to another molecule. To give a simple example, imagine that you mix purified EGF with its receptor, EGFR. If you run this through an analysis you can see EGF in two peaks: one peak will be free EGF and the other peak will be EGF bound to EGFR. So technically, you can have an analyte present in multiple peaks, and some methods may do a better job of separating those peaks than others (ie. you might see a single peak in HPLC but multiple in CE as per your example).

However, in this case I think that is very unlikely. They're working with a mixture of carbohydrates, and these don't really interact with each other at all. So through a CE or HPLC run, an individual carbohydrate will be present in one peak and one peak only (provided you've optimized your separation conditions).

2) Is there any way to do NMR on an unseparated mixture and see the signature of a particular component in an identifiable way? (Amphastars claim is that it is not possible because all the different components of the mixture interfere with each other - but I doubt the veracity of Amphastar's argument because, for instance, spectrographs accomplish such measurements all the time on heterogeneous mixtures.)

I'm not super strong on NMR, but I suspect Amphastar is taking the lazy way out here. The main reason I say that is because NMR for structures like carbohydrates is, to my knowledge, relatively straightforward*. So you can create mixtures and run the NMR and see what the mixture's fingerprint looks like, and you can identify the NMR peaks and spots since you know exactly what you put in there**. With these fingerprints then, you should be able to decipher constituents in complex mixtures.



* I'm sure there are complexities, but I think this is reliably true. Especially when compared to NMR for proteins.

** This includes labeled methods using 15N, 13C to provide certainty. For example, if you suspect peak A is a mixture of compound X, Y and Z, then you can prepare separate, labeled compounds X*, Y* and Z* and then rerun the NMR with various setups (X*, Y, Z; X, Y*, Z; X, Y, Z*) to deduce which signal in the mixture is from which compound.