Alkaloid identity and mass spectrometry

[0_o]

I searched for this topic but could not find it (if it has been addressed) so...
The following reference is primary to this topic:

Corio C, Soto IM, Carreira VP, Padró J, Betti MIL, et al. (2013) An alkaloid fraction extracted from the cactus Trichocereus terschekii affects fitness components in the cactophilic fly Drosophila buzzatii. . Biol J Linn Soc Lond. 109 (2): 342–353

In this study the authors were looking at the alkaloids of Trichocereus tercheckii and they observed mescaline(3,4,5-trimethoxyphenylethylamine), as well as trichocereine (N,N dimethyl-3,4,5-trimethoxyphenylethylamine) and another alkaloid. In this study they climed to be able to identify this alkaloid via reported values for GCMS, no confirmation with a reference molecule was made. They identified the alkaloid as α-methylmescaline, that is mescaline with a methyl group (CH3) on the α carbon, α-methylphenylethylamine is known as amphetamine, and 3,4,5-trimethoxyamphetamine one of the 6 isomers of TMA (there are 2 enantiomers for each of the 3 possible TMAs) is the amphetamine equivalent of mescaline.

In Differences in Tolerance to Host Cactus Alkaloids in
Drosophila koepferae and D. buzzatii the same approach was employed again:

"The identification of the active components in the alkaloid fraction was accomplished via mass spectrometry. We confirmed the presence of three compounds: trichocereine, N-dimethylmescaline, a phenylethylamine alkaloid typical of this species, mescaline and the analogue α-methylmescaline"

Note that they reported a molecule with two methyl groups at the amine group, as well as the parent molecule lacking the methyl groups at the amine but did not report an intermediate.
Note that α-methylmescaline had never previously been reported from cacti, moreover no cactus alkaloids based on phenylethylamine with alpha methylation had been reported in cacti.

What had actually happened was that the studies involved did indeed employ mass spectrometry, however that is not a means of identification it is solely a means of indication. The authors mistakenly believed that they could rely upon published values for GCMS to identify what alkaloids were present. Alpha-methyl-mescaline and N-methyl-mescaline have the same molecular weight and formula, and while their breakdown products are slightly different the authors failed to confirm the indication of the molecule that GCMS provides with an identity confirming test such as running a pure reference sample of the suspected molecule. Nor did they appear to investigate which enatiomer of their suspected molecule was present.

The point here is that that GCMS is a valuable tool, however it is not a means of identification and replying upon published values to identify a molecule is not a valid means of identification, rather it is a viable means of indication requiring further confirmation.

I raise this issue as that the practice of using GCMS data without a confirmation is becoming common in academia and other places and it should be known that it is less than ideal as that it has and will lead to misidentification at times and this will subsequently result in misinformation.

 

[benzyme]

no kidding.

there is a reason compounds like THC-A aren't analyzed using GCMS. LCMS is more versatile, due to its ability to analyze heat-labile compounds

 

[0_o]

The issue is not merely limited to GCMS, the problem is the use of published chemical data for mass spectroscopy as a means of identification when it is only a means of indication.

There are also two extensions of the topic in terms of information and impact.

The first is that claims are being made not only about identification without any confirmation tests such as excitation and emission, melting point etc... claims are being made against the presence of molecules based on published peak data for mass spectroscopy. When a molecule is found but TLC and or mass spectroscopy raise issues with identity then claims are being made as to what the molecule is not, however isolation and testing of the molecule are not being made. Plant extracts and artifacts can complicate spectroscopic and chromatographic results and this results in the potential for misinformation to arise not only in terms of claims of what molecules are but also in terms of what molecules aren't.

The second extension of the issue in terms of effect or impact is that the same issues that are complicating findings and claims are not limited to academic and illicit investigation and publication... rather these issues are also found in the field of forensic science where they impact prosecution. People can lose their liberty and property as a result of inconclusive and uncertain claims being presented in court as conclusive and certain by so called experts. The lack of objective and adequate peer review that is common in publication is essentially systemic and ubiquitous in forensics.

The use of published spectral and breakdown peak data is presented as certain and conclusive when it is not, this issue does not merely create misconceptions about what alkaloids are and are not present in various plants and materials... it actually has a profound negative impact upon the lives and liberty of numerous individuals through so called forensic evidence.

 

[benzyme]

you're preaching to the choir. My employer performs such testing, but ultimately, legal money interprets the results. Oligarchs, whether it be in the form of publishers or legal review boards, dictate what gets presented and interpreted,not the researchers.

btw..mass spectrometry can absolutely be used as a means of identification, using either data-dependent or data-independent methods; it requires subscription to databases like NIST and/or Wiley. Tandem mode analysis with collision-induced dissociation is particularly effective at elucidation of isomers/stereoisomers and epimers, especially when separation is done with complexation with crown ethers or cyclodextrins.

 

[0_o]

It is my personal conviction that mass spectrometry is not sufficient to result in an identification, only indication with degrees of probability.

The indication may be to a high degree of probability, however without confirmation employing isolation and characterization it is my belief that identification has not been made, only indicated at best.

In my own reviews of literature I find that the NIST database is a useful tool but provides mere indication not identification and that misidentification and subsequently misinformation have resulted in some cases. Without characterizing an isolated chemical or compound I cannot view any claims arising from database correlations as certain.

 

[benzyme]

what you described can be applied to all science...because nothing in science is proven. The method can only show probabilities of correlation.

most papers which characterize compounds use multiple methods; and despite your belief, it is widely accepted in academia and industry, that MS, and especially MS/MS, is a definitive means of identification of compounds. In the case of unknowns, preliminary investigation, as with TLC, may constitute indication. This is why standards/controls are used...for confirmation/identification.

 

[0_o]

it is widely accepted in academia and industry, that MS, and especially MS/MS, is a definitive means of identification of compounds.

This is why I began this discussion with evidence against it.
The practice of using the NIST library as a means of identification results in misinformation and misidentification in too many cases to ignore.

I also disagree that nothing is ever proven.

If you have a video of a duck, that proves it is a duck.

But if instead you cut the duck into pieces and weigh them and check NIST for what has parts that weigh that much... its silly, yet this is what is currently accepted.

 

[benzyme]

Show me a paper that "proves" something, and I'll show you "researchers" trying to sell you something.
Scientists never use "proven" in a discussion, there is only evidence to show correlation, or lack thereof. You'd know this if you ever published anything. R^2 never equals 1.

"There is strong evidence to suggest that it may be a duck."

as far as using NIST..it uses an algorithm to show probability, like PROWL, Seaquest, and other bioinformatics
databases. Sure, there is an error factor, like with anything else in science. No instrument/system is perfect.

 

[0_o]

https://link.springer.com/content/pdf/10.1007%2Fs13361-016-1589-4.pdf

^
A good solution to part of the issue.

As for proof, it really isn't worth debating.
Many things are well proven.
Take for example the periodic table, atomic weights, functional groups etc.
The idea that nothing is proven seems silly to me and I won't bother addressing it further.

 

[Aum_Shanti]

Take for example the periodic table, atomic weights, functional groups etc.
The idea that nothing is proven seems silly to me and I won't bother addressing it further.

These things are also not PROVEN.

It's just that all current experimental results match these abstracted theories and therefore they get accepted as current best explanation and easy to work with abstractions to be of use. But there could potentially always come an effect around the corner which would need an adjustment or a complete rejection of current theories. This happened in the past, and could well happen again in the future...

 

[endlessness]

You all have a point even though it seems you are arguing against each other.

0_o , you seem to be saying that merely depending on NIST database for identification of a compound is insufficient specially in something like court cases. There are several compounds which may have similar mass spectra and specially when dealing with low signal, it makes it less reliable. That being said, benzyme clearly mentioned that when it is a more serious subject, the good scientists will not simply depend on database ID but will run side-by-side with a reference standard. In that case, you won't have only the database ID but you will confirm with your reference standard which will be at the exact same elution time, plus the mass spectra comparison. Also, in my understanding, MS/MS will make it even more reliable since further breakdown of your compound will make it even harder to mistake it for another substance.

Without a reference standard I agree it should be called a mere indication. With a reference standard I think you can already be pretty sure. Of course if you want to be 100% sure and someone's life depend on it or it is about to be part of a major publication with groundbreaking results, then further confirmation with NMR, x-ray crystallography or other complementary methods is always ideal.

As for the "proven" discussion, it is relevant to bring into the discussion what Benz and Aum_Shanti said, that we only work with models, and that even if all data is consistent with our model, it doesn't mean there won't be in the future some contradicting data that will require us to change our model. This is why nothing is "proven" in that sense.

Yet, if we have that in mind, I think it's ok to use the word "proven", because, after all, it's a word that exists, and it has its function. It serves to make a distinction between something that has very high reliability, and something that doesn't.

Take for example, airplanes.. We can say it is "proven" airplanes can fly, or that it is proven the shape of the airplane wings will, when at a certain speed, move the air towards the ground creating a lift. Sure you can invoke the "but nothing is proven and we may change our model" argument, but that affirmation is proven enough that you even can trust your life on it, right? Millions of people do so daily when boarding an airplane. So to argue that "nothing is proven" given that affirmation is more like, arguing semantics, which is fine I guess, but if we all already understand that all models are subject to revision and won't close our eyes if new evidence come in, it is unnecessary to argue those semantics too much.

And just to bring back to the discussion using the above commentary as a guide, you can say it is proven (enough)what the identity of a compound is when you have LC-MS/MS, H-NMR, C-NMR and x-ray cristalography and a reference standard next to it all agreeing on that ID. Can we call it "proven" when we have just GC/LC-MS and NIST database? I think most of us would agree that not. Can we call it "proven" if it's GC/LC-MS but not just a database, rather comparing with an authentic reference standard? That is where probably it could be argued differently and I'm not sure what the above posters think.

 

[Loveall]

I think you can prove theorems inside an abstract axiom framework. If you consider math a science (and not a branch of philosophy), then I guess you can prove stuff in science. There are theorems that have survived over millennia (e.g. the Pythagorean theorem).

On the other hand, I don't think anything in nature and natural science can be proven. For example, the duck in the video could be a robot. Scientific theories (or models) are usually revised, refined, and updated.

 

[0_o]

You all have a point even though it seems you are arguing against each other.

Agreed.

Without a reference standard I agree it should be called a mere indication. With a reference standard I think you can already be pretty sure.

Agreed.

As for the "proven" discussion, it is relevant to bring into the discussion what Benz and Aum_Shanti said, that we only work with models, and that even if all data is consistent with our model, it doesn't mean there won't be in the future some contradicting data that will require us to change our model. This is why nothing is "proven" in that sense.

Agreed.

Yet, if we have that in mind, I think it's ok to use the word "proven", because, after all, it's a word that exists, and it has its function. It serves to make a distinction between something that has very high reliability, and something that doesn't.

Agreed.

I submit my own belief is that after something has been repeatedly tested and measured and has withstood all attempts to falsify it then that thing (so to speak) is proven until such a time as it is dis-proven.

If however we maintain that nothing is ever proven (to be X), then we can also maintain that nothing is ever dis-proven (proven to not be X). Moreover if nothing is ever proven then the idea that nothing is ever proven cannot be proven and thus becomes absurd, it is a self contradiction. If the statement; nothing is ever proven is true... something has been proven and the statement is not true.

As for the airplane... I maintain that they are proven to be able to fly, but that in individual cases they may dis-prove that by falling out of the sky. However it is true there are those who maintain that humans create reality and that we can alter it on purpose and if that is true then one day all of the airplanes will fall out of the sky and no longer fly if someone changes reality to that. Perhaps there is a dimension where right now all planes are falling out of the sky for just that reason? I cannot disprove that, however i have no evidence for it and so can dismiss it as not being something I can personally believe.

I should explain that it is my observation, experience and belief that things can be proven and dis-proven irrespective of belief and dis-belief.

However I am seeking to support a conclusion (via statements and evidences) that mass spectroscopy and library reference (such as NIST) is insufficient to prove or disprove what a chemical is or isn't. I do not dismiss Mass Spectroscopy nor data libraries as a tool.

 

[0_o]

On the other hand, I don't think anything in nature and natural science can be proven. For example, the duck in the video could be a robot. Scientific theories (or models) are usually revised, refined, and updated.

True enough.
A video would not be enough to prove the duck.

If we look up the weight of a witch and the duck does not weighs the same we could then publish that we know that it isn't a duck.

 

[Mindlusion]

It is my personal conviction that mass spectrometry is not sufficient to result in an identification, only indication with degrees of probability.

The indication may be to a high degree of probability, however without confirmation employing isolation and characterization it is my belief that identification has not been made, only indicated at best.

In my own reviews of literature, I find that the NIST database is a useful tool but provides mere indication not identification and that misidentification and subsequently misinformation have resulted in some cases. Without characterizing an isolated chemical or compound I cannot view any claims arising from database correlations as certain.

High res mass spec by means of orbitrap can give you an accurate mass with high precision down to 5 decimal places. With lock mass correction it's relatively easy to keep that precision even without regular internal calibration.
With this kind of accuracy, you can resolve peaks of isotopic compositions of the same mass, between two hydrogens and a deuterium, or between a molecule which contains a C13 atom or the same molecule containing a deuterium atom.
At this degree of accuracy, the mass of the electron is taken into account (0.00055 u)

for example, lets take the molecule LSD
It has the molecular formula of C20H25N3O

and a theoretical monoisoptopic mass (exact mass) of 323.19976 u



lets assume the worst and drop the 5th decimal place in the measurement, the orbitrap gives us 323.1997


which means its within 0.00006 u (0.06 mu) of the only possible molecular formula with the mass of 323.19976


using this exact mass to formula calculator here
http://www.chemcalc.org/mf_finder/mfFinder_em_new


there are countless other formulas with the mass of 323.2, the next closest formula C15H24FN6O
which has a mass of 323.199623. which is a whopping 0.137 mu distance from the exact mass of LSD. A full power greater or 23x further discrepancy. In practice the accuracy is even better, usually giving it right on the dot at 5 decimal places or within 0.00003 u.

And now if you apply common sense and realize that the formula C15H24FN6O containing a fluorine and 6 nitrogens, does not exist in the material you are analyzing because it doens't exist in nature, and that there are no known stable structures for such a molecule. Then your certainty just goes up.

In the end, the chances of misidentifying a molecular structure in a plant sample with another possible structure in the same sample is virtually zero. You can even tell apart structural isomers by collecting a fingerprint to near the same degree of certainty using collisionally induced dissociation as a soft fragmentation technique. The collected fragmentation pattern is essentially a fingerprint of a structure, with soft ionization techniques you can vary the energy to collect a fingerprint across a gradient, giving you even more data points unique to that structure.
Fingerprinting is done analogously with simpler techniques like IR and Raman, which are used extensively in the industry.

its not a matter of opinion, its science. Its accuracy and precision. Sure you can say it doesn't
'prove' anything, nothing truly can, rationality itself cannot be proven by its own logic. If it gives the correct result a billion times in a row, i wont live long enough to see a billion so it might as well be 100% accurate. its damn useful technology.

In terms of elucidating the structure of an unknown, of course, while mass spec is still extremely helpful, isolation and further examination by NMR techniques and x-ray crystallography are necessary to determine exact stereochemistry.

But even if you have a vague idea of what the structure should be, you're in the clear of within 10^5 degrees of accuracy

 

[0_o]

Good points.
My point is only that misidentification occurs, human error being one of the factors, and it is documented.

Hence actual confirmation is needed.

Maybe one day errors and misidentification will no longer occur.