Which protons are exchanging that give us the peaks?

[srveale]

Here's an arginine NMR spectrum. It doesn't say what the reference is so I'm assuming it's TMS.

Here's a 3D viewer for the molecule:

http://bmrb.wisc.edu/metabolomics/mol_summary/jmol_display.php?path=/metabolomics/standards/L_arginine/lit/3362.mol&mol=L-Arginine&bmrbid=bmse000029#

And here's which atoms they assigned to which peak (there's an option in the 3D viewer to show atom ID):

"Assigned Chemical shifts for bmse000029, L-Arginine"

Shift ID Atom ID Chemical shift Ambiguity 1 C2 57.002 1 2 C3 177.238 1 3 C5 30.281 1 4 C6 26.577 1 5 C7 43.201 1 6 C9 159.504 1 7 H15 3.764 1 8 H16 1.909 1 9 H17 1.909 1 10 H18 1.679 1 11 H19 1.679 1 12 H20 3.236 1 13 H21 3.236 1

So my guess is we are interested in the main chain hydrogens, ie H16 and H17, and the second peak that appears at low pH presumably corresponds to H18 and H19. Is this the right direction?

[DrSAR]

Agreed that this is TMS-referenced. Further, I believe this is arginine in water which should give rise for a huge water peak (50M vs maybe 1mM?) were it not for solvent suppression in the sequence use. That solvent suppression gives rise to that slightly odd lineshape for the water peak which comes at around 4.9ppm (but isn't always there can be as low as 4.5ppm).

The protons you refer to H16-H19 are all around 1.7-1.9ppm and if what's left of my chemistry knowledge doesnt deceive me are all 'aliphatic'. In our referencing scheme these would be appearing at 1.8-4.9=-3.1ppm. Ring a bell? Given that our peaks are at 3ppm I worry slightly. Did you guys get the sign for the offsets right? Where does the other peak we sometimes see land on that spectrum you show?

Also, peaks do shift somewhat depending on pH and pH3 is a bit extreme compared to what might have been used for the predicted peaks you list.

[firasm]

The other peak is at 1.9 ppm, here is an image of the peaks at pH 2.

[srveale]

My mistake, the reference is listed as DSS, and the solvent is D2O

http://bmrb.wisc.edu/metabolomics/mol_summary/show_data.php?molName=L_arginine&id=bmse000029&whichTab=1

[DrSAR]

OK, DSS gives virtually the same shift as TMS. Also, I believe that 100%D2O are more like 99% or so which means that there is probably still a lot more H2O than DSS or sample in the NMR tube (unless you go to 100mM). Hence we should still see a water peak at 4.7ppm or so.

I still worry about our peaks at 3ppm and 2ppm since they would be 7.7ppm and 6.7ppm with respect to TMS (or DSS), respectively. Do I have this wrong?

[srveale]

Our peaks are ~3.1ppm and ~1.9ppm. If we... pretend these are negative, it matches up with those groups of peaks in the NMR spectrum.

Could it be that the lab computer knows about the reverse convention, and gives ppm values as negative for greater frequencies? So in effect we've been flipping the image 180 degrees, twice?

[firasm]

I think @erinmacmillan's expertise may prove useful here.

In the meantime, Bard & Ballaban report the arginine peaks to be at 3 ppm and 2 ppm respectively, that's consistent with what we see, so something is off here.

The plot thickens...

I've also sent a snippet of this to a friend in NMR spectroscopy as well.

[DrSAR]

The resonance freq of a hydrogen species is an absolute thing. Only because it's tedious to talk about a peak having a frequency of 300.00012345MHz do we use relative frequencies. In NMR if you were to reference the resonance of the aliphatic peaks with respect to water rather than TMS, I am fairly sure you would get -3ppm (ish).

Here is my discussion of some possible explanations/excuses:

• With regards to the console making mistakes on our behalf: All we do is program freq offsets that result in the sat pulse being played out at bf0+offset. So there is no inversion there (unless we did something truly funny in the methods code).
• The Bard & Ballaban reference @firasm is showing is talking about -NH2. That would be protons H13, H14, and H22-H25 in the BMRB list. None of them have freq predictions.
• There could be a convention of the z-spectral people such that the frequency for z spectra is reversed. Why would anyone choose that?

Taking bets now. Mine is on us talking about different peaks. In other words I don't think we see the CH2 protons.

[erinmacmillan]

Hi, fun discussion!

I definitely agree that the peaks you see at ~6.6 and ~7.8 ppm are not on the BMRB list pasted above, and arise from NH protons. Also, due to exchange and differences in pH and temperature, these peaks will appear at different frequencies in different experimental conditions. The best spectra of these exchangeable peaks I could find in less than 10 min of google searching are from Klevan and Crothers, High-Resolution NMR of Exchangeable Protons in Arginine, Oligoarginines, and the Arginine-Rich Histone Tetramer, Biopolymers, 1979;18(5):1029-1044

These spectra were acquired in H2O, referenced to DSS, at pH 5.3 and increasing temperature. As you can see, these peaks experience exchange-induced frequency shifts, line broadening, frequency averaging. You can find more details about these phenomena in the Levitt book chapter 19.

Klevan and Crothers assign the 8 ppm peak to the alpha amino group (see structure below), the 7.44 ppm peak to the N1 proton, and the two peaks at the lowest ppm to protons on the two N2 at the end.

So, I'm not sure which of these you are seeing at the pH and temperature you used (pH 3 and room temp?). Maybe worth making a pH 5.3 phantom to compare with these spectra?

Hope this helps! Erin

[DrSAR]

Thanks Erin - this paper from '79 (who around here was born then?) is a very good find. I suggest we all read it in great detail since it's pointing to a few features quite relevant to CEST MRI. Not least of which is Fig 6(a) with a computer (!) simulation for the coalescence of two resonances as a function of increasing exchange rate.

[DrSAR]

At different pH rates and our current choice of sequence parameter (saturation pulse etc), these peaks disappear quite quickly. @firasm or @srveale could possibly post a stack of z-spectra with different pH values (all at room T, I believe). An experiment that might get us some data to compare to Fig 2 from Klevan and friends that @erinmacmillan posted, would be to look at arginine near 0C and then again at room T. This doesn't make publication worthy graphs but it is a good educational exercise.

[srveale]

Thanks Erin for the help! These are the spectra at room temp. The one we've been talking about is at the lowest pH of about 2, where the second peak is largest. We don't see it above pH ~3.