Chromatography Forum

I was recently speaking to a number of very experienced (certainly compared to me anyhow) practitioners in metabolite identification. They were from 4 different pharmas but they all basically gave the same response. Their goal is to rapidly identify "active" metabolites and they all seemed to use basically the same approach. My argument was they were barking up the wrong tree as they could not guarantee that their strategies would pick up either all of the actives or indeed any. I suggested a better approach was to be able to at least identify all of the metabolites produced and then to use this as a starting point for further work in identifying any "actives". My feeling being that at least they wouldnt miss anything with this approach. Unfortunately it was neither the time nor the place for a heated debate but I would very much like to hear the opinions of the members of the forum.

I am not familar in this field, but very interested in it. Would you suggest some learning resources? Thanks a million.

I don't know what types of metabolites you want to look at, but if you've screened any soups, you will know there is a huge forest of compounds from most, unless they are based on "defined media", when it's more of a medium sized forest . It's easy to change the metabolite profile by stressing the organism by enviroment, including media, giving even more diversity.

The marine organism metabolite enviroment is probably the most unexplored screening territory, and should keep many people employed until retirement.

Generally, there are some initial parameters, such as the molecule must be reasonably stable, have the desired properties ( eg if looking for cytotoxins, you would like it to be cytotoxic ), and the source and growth is able to be easily replicated for the first few tests. No point in working on a molecule that requires huge investment to manipulate it further, if you can't be sure there's a market.

Screening is inherently about throwing out 99%, and focusing on the 1% that might match your criteria. The limitation may be more about the biological activity ( which could even be synergistic from several molecules ).

I'm not sure what environment you are in, but in some drug discovery, activity is first tested biologically, and then positives are crudely chromatographed, and the fractions both characterised and retested for the biological activities of interest. MS libraries are then used to separate knowns and unknowns, and further characterisation performed on unknowns of interest.

The idea of identifying all molecular metabolites, regardless of utility is probably very resource intensive, and would be hard to justify to most bean counters holding the purse. Targetted searching probably seems intuitively sensible to them.

Bruce Hamilton

Bruce, with agricultural chemicals (where the concern is the metabolites produced by the myriad soil organisms), a standard approach is to incubate with radio-labelled active. You have to know enough about the molecule and its potential metabolic/degradation pathways to put the hot atoms in the right place(s), but it does provide a powerful way of of finding out what parts of the initial screening runs you can ignore.

This is not a good idea with human subjects, however

Oops. Thanks Tom.

I grabbed the wrong end of the stick!. Sorry about that.

I did wonder whether "active metabolites" referred to known chemicals, but I assumed modelling would cover concerns about missing possible pathways. I analyse some labelled metabolites and drugs, so I should have thought about that possibility more.

I sailed on regardless, because I'm more familiar with the term in drug discovery, looking for new drugs from organisms. Most of the actives of pharmaceutical interest are mebolites from yeasts etc..

Hope I didn't confuse anybody else.

Bruce Hamilton

Tom, many of the metabolic pathways in humans have been elucidated that way.

Maybe I was a bit vague as this isnt really my area I just got chatting to these guys. What they were all doing was taking a drug of interest and putting it through a screen or an animal and looking for the metabolites produced. Of course sometimes the metabolites are also active and hence are of interest as they could potentially be other drugs or a particular pathway may need to be inhibited.

My argument with them was they couldnt be 100% (probably no better than 70%) sure that by adding a compound to react in a certain way they would be able to spot an active metabolite. And doing it this way they could only look at one or two chemistries simultaneously and so could still be missing actives

Adding a marker such as a radio label seemed to me to be a more obvious way of spotting ALL metabolites and this would be a better starting point. I understand the problems of radiolabels in terms of dose and cost but as a general principle I thought it was more sound.

I hate to be critical but it struck me they were all set in their way of doing things and were not looking for better ways. I could be totally wrong but I was just genuinely interested to hear any other points of view.

Tom, many of the metabolic pathways in humans have been elucidated that way.

HW, as you can tell, this is out of my area but I would imagine that it would be a lot easier to use radio-labelled compounds to discover metabolites in tissue culture versus live subjects.

Just an example: About 30 years ago it became important to know how much glucose the brain burns in relation to the rest of the body. This was easily determined by people eating some C-14 glucose. At rest it was about 50%, when the person moved the % utilized by the brain (the absolute amount stayed about constant) dropped to below 5%. There are zillion problems unsolved because people are affraid of radioactivity, yet they readily succumb to X-ray diagnosis (rough guess: ~ a million times more damage done). In nuclear med diagnosis (which is sort of an in vivo metabolic study) or therapy one uses doses that are also much higher than necessary for in vitro stuff.