Quantitative analysis of known compounds!

[MS-NMR]

Hello everybody,

I’m a newbie in the field of GC/MS. I’m doing quantitative GC/MS analysis of known weak acids such as fumaric acid, succinic acid, and some amino acids in cell extracts but I have no idea how I can do the quantification! I incubate the cells with [U-13C]glucose and after incubation the weak acids and Amino acids will be 13C-labelled thus I think I can’t use inter standard.
I’d be very grateful if someone tells me what can I do?

Best Regards,

MS-NMR

[Peter Apps]

Are you interested only in the labelled acids, the labelled and unlabelled seperately, or the total of the two together ?

In their native form these acids are not amenable to GC, what sample prep are you doing ?

Peter

[MS-NMR]

Are you interested only in the labelled acids, the labelled and unlabelled seperately, or the total of the two together ?

In their native form these acids are not amenable to GC, what sample prep are you doing ?

Peter

Dear Peter,

Thank you. In fact I'm interested in new synthesized acids from 13C-labelled glucose. I derivatize the samples using MTBSTFA.

[Don_Hilton]

The biggest problem you have with quantification of the acids is that you have signals for the various isotope mixtures all mixed together. You may be able to quantify the total for a particular acid by summing all ions togther in the isitope cluster for the molecuar ion and assuming the same response factor as for the sumed isitope cluster for the native material. To quantiy the specific levels of inclusion, you have another challange. You can calculate isotope ratios for the lableled compounds and try treat the spectrum of the molecuar ion isotope cluster as the sum of multiple spectra - and use algebra to attempt to detemine the rate of inclusion of 13C. Both of these sugestions are based on significant assumptions and would need to be checked out carefully - but if it helps, let me know.

If summing signals works, an internal standard is optional. The problem is in selecting a signal for quantifying compunds with varying isotope mixtures.

[MS-NMR]

The biggest problem you have with quantification of the acids is that you have signals for the various isotope mixtures all mixed together. You may be able to quantify the total for a particular acid by summing all ions togther in the isitope cluster for the molecuar ion and assuming the same response factor as for the sumed isitope cluster for the native material. To quantiy the specific levels of inclusion, you have another challange. You can calculate isotope ratios for the lableled compounds and try treat the spectrum of the molecuar ion isotope cluster as the sum of multiple spectra - and use algebra to attempt to detemine the rate of inclusion of 13C. Both of these sugestions are based on significant assumptions and would need to be checked out carefully - but if it helps, let me know.

If summing signals works, an internal standard is optional. The problem is in selecting a signal for quantifying compunds with varying isotope mixtures.

Dear Don_Hilton,

Thank you for your recommendation. Unfortunately I can’t use summing signals of a particular acid because the acids can be produced via various ways in cells. In fact I will study the effect of (for example a drug) on these ways. Therefore each signal of a particular acid must be separately quantified. My problem is not isotope mixtures but I don’t know how I can quantify the isotopes.
I think it helps me if:
1) I incubate the cells with 13C-labellet tracer.
2) I use another cells without incubation with tracer as bulk.
3) I make a mix from various acids as standard.
I reduce the value of exp. 1 from exp.2 and using exp.3 I make for each acid a standard curve and then I determine the concentration of each isotope of a particular acid using the standard curve of the same acid. I’d like to ask you whether am I correct with this method?

Thank you

[Peter Apps]

This is going to be difficult - has it ever been done before ? Your first problem is to generate a signal for each of the differently labelled molecules that you are interested in - once you have done that the quantification will be reasonably straightforward since labelling should have a only a small effect on fragmentation.

Nominally the mass spectometer can discriminate between molecules with different numbers of labelled carbons, but with the limitation that it cannot tell the difference between carbon isotopes and isotopes of other elements. A non-polar column with sufficient resolution can separate molecules with different degrees of isotope labelling, but unless you use very long, narrow bore columns a single C13 label will give only a broadening of the peak for the unlabelled molecule.

A high res MS could discriminate the labelled molecules without interference from isotopes of other elements.

Peter

[HW Mueller]

Certainly, this type of thing has been done many times, most likely much less ambitious though. I once helped a doctoral student to quantify C-13 inclusion in oxalic acid from urine after administration of C-13 ascorbic acid. Now this was a simple problem compared to what is at hand here, still the GC-MS of methyl esters of the acid would have had to be optimized considerably more. The student didn´t help (not atypical in medicine here) so I dropped this project, though I had spent consideable time in the workup. (This was a long time ago so the details have escaped me, but I think I isolated the oxalic with HPLC, esterified, etc.).
Anyway, depending on what MS-NMR really wants to do, he could have a life project at hand: Qualitatively determine all the C-13 pösitions in the products, etc. etc.
Just remember: Helped another student with C-14 CO2 inclusion in ascorbic acid in tomatoes which where grown with (14C)CO2 in the atmosphere. Here the student gave up, because we couldn´t detect any C-14 in the HPLC´ed ascorbic. It turned out that the "team" of the main advisor and student had used way too little labeled CO2 in the atmosphere.
I am mentioning these examples to show that some good professional(s) need to back up such experiments. (Incidentally, I had advised not to check into the position of C-14 as I expected no separation of ascorbic acid with C-14 in different positions with the HPLC).

[Don_Hilton]

With GC/MS you are only going to be able to determine the total inclusion of 13C in a compund - at best. You need to obtain a good separation of each acid from the others to get a clean enough signal to do that. If you need to determine the location of the included atoms, you will have to isolate the acids and use some other technique.

The separating capabilty of GC/MS helps in that you can separate out individual acids. (And that separation is required for the suggestion I gave above.) The fact that the chemistry of 12C is so nearly indistinguishable from that of 13C keeps you from being able to see separation of labeled molecues from the native form in chromatographic techniques - keeping the signals mixed together. The lack of any chromatographic sepration does aid in the use of mathematical separation of the signals.

The mixed standard you suggest has a level of complexity that works agasint you. If we have a four carbon acid, the acid can contain one, two, three, or four 13C atoms. For each mixture of isotopes, the spectrum will be different and the position of the most intense ion in the isotope cluster will change. The natural presence of 18O will add a bit to the complexity of the spectra.

[Peter Apps]

The increment in molecular weight that comes from isotope labelling can be enough to chromatographically separate labelled from unlabelled molecules, and those with different levels of labelling. You can enrich uranium that way.

Peter

[MS-NMR]

The increment in molecular weight that comes from isotope labelling can be enough to chromatographically separate labelled from unlabelled molecules, and those with different levels of labelling. You can enrich uranium that way.

Peter

If it the separation of isotopes of a particular acid was not possible, can I use the extracted ion chromatogram for each isotope?

[HW Mueller]

What is "extracted ion chromatogram"?
If the glucose has only one C-13 in it it is unlikely that the acids will have more than one.
On separating isotopes: A colleague separated sugars with D and H routinely with GC.

[Peter Apps]

If it the separation of isotopes of a particular acid was not possible, can I use the extracted ion chromatogram for each isotope?

You can, but remember that things other than a labelled carbon can give you an MS peak one unit heavier than the peak from the unlabelled molecule. Keep in mind also that labelled carbons will not be in every fragment of the spectrum.

Are you expecting only one labelled carbon in each acid ?

Peter

[MS-NMR]

If it the separation of isotopes of a particular acid was not possible, can I use the extracted ion chromatogram for each isotope?

You can, but remember that things other than a labelled carbon can give you an MS peak one unit heavier than the peak from the unlabelled molecule. Keep in mind also that labelled carbons will not be in every fragment of the spectrum.

Are you expecting only one labelled carbon in each acid ?

Peter

No. Maybe 2 or 3 or more labelled carbon are in each acid. For example if I use [U-13C]glucose as tracer, after first turn of TCA cycle, the following labelled glutamate (glu) can be produced.
[4,5-13C]glu, [1,2,3-13C]glu, [2,3-13C]glu, [2,3,4,5-13C]glu or [U-13C]glu.

[MS-NMR]

What is "extracted ion chromatogram"?
If the glucose has only one C-13 in it it is unlikely that the acids will have more than one.
On separating isotopes: A colleague separated sugars with D and H routinely with GC.

Dear HW Mueller,

In my case each chromatogram includes some Ions such [M-57]+, [M-189]+. I can using processing software divide each chromatogram in smaller parts.
I will use [U-13C]glucose as tracer.

[Don_Hilton]

The separation of deuterium labeled compunds by GC works well, but 13C labeled compounds do not separate well from the natives. In our laboratory we analyze native compunds and use internal standards with multiple 13C atoms in them. Whlie the isotopically labeled compounds may not elute quite exactly with the natives - it is not enough separation to notice - and I look for one peak marker through the native and the labeled comounds.

On the inclusion of 13C in a molecule, remember that the native compunds have 13C already in them. Assuming derivitization: The TMS ether of a four carbon diacid will show 10% relative abundance because the natural 13C present. (A total of ten carbon atoms - 3 in each TMS and four in the acid.) Thus, to see inclusion, you need to have sufficient 13C in the glucose source to ensure that the acids you analyze are formed from the added glucose. And if you use a TMS ester, the silicon contributes to the m+1 abundance as well.

A dimethyl ester of a our carbon diacid would show about 6% relative abundance for m+1.

I would suggest running a mixture of acids down the GC and make sure you have then well seprated and if you can get some labeled acids, make some mixtures and shoot them to see what the combined spectra would look like. You will need some monolabeld, dilabeled, etc because the spectra for each will be different.