«

»

Aug
16

Finding invisible neglected TB in vivo active compounds with ChemSpider

Very occasionally I come across molecules that cause me to scratch my head and do a double take. I frequently scour databases to find obscure ones to build structure activity relationships or try to learn from the past. Sometimes you uncover surprises along the way – or in one case, many .

24hrs ago I was digging through a paper from 1987 on “Aspects of the antituberculosis activity of 27753-RP, a new semisynthetic derivative of griselimycene” by M. Toyohara. After testing in mice it was suggested that  27753-RP was a highly promising antituberculosis drug. In an unsupported statement the author also stated that  griselimycine while effective in vitro against TB was ineffective in vivo. So I looked for more such papers on this compounds and could find nothing. So I have no idea what happened to it. I was curious to find the structures of both molecules.

First my go-to database, ChemSpider could not find a match for 27753-RP  then the match for griselimycene was unusual, it was too small and certainly not a large macrocycle. I pretty much knew this was going to be interesting. I quickly popped an email to the folks at ChemSpider and the this morning the synonym “griselimycene” was removed from the small molecule. In the interim I managed to find how the same synonym error was in several other places on the web here, here, and here. 

So you can still see what I saw yesterday, a musk compound with the wrong synonym. Some digging by Antony Williams from the RSC demonstrated how the incorrect molecule had spread around since being deposited in PubChem..and so this provides another example of how things can spread as reported earlier.

Tony was also able to find griselimycin – notice the lack of (e) – perhaps a French spelling? Not all the stereocenters were defined (9 of 12) and there was a second version with 11 of 15 defined. Finding 27753-RP proved another challenge. In the paper the author incorrectly cited the wrong molecular formula C67 H11 N10 O11 – notice the very low number of hydrogens compared with griselimycin C57H96N10O12. Searching ChemSpider for C67 H116 N10 O11 retrieved the compound and in the synonyms we found 27753R.P.  So the periods made all the difference in making this compound invisible to the original searches. Tony edited the record to include the correct synonym from the paper. Again no stereocenters are defined for the structure.

This just goes to show how difficult it can be to find molecules. Now we have them we can see that the addition of a long aliphatic chain on the molecule maybe the difference between an in vivo active compound and one that is inactive. Of course there is a shift in ACD logP from 2.97 (griselimycin) to 8.56 (27753-RP) which could be important too. Perhaps the compound is more metabolically stable or has better bioavailability or just gets into TB better.

Maybe this compound deserves further characterization because it was active against a rifampicin resistant strain of TB. How many more examples exist like this which have been overlooked, reside in obscure journals or their structures are virtually impossible to find? Thanks to Antony Williams, David Sharpe and colleagues at ChemSpider I think there is a chance we just might find more like this!

 

 

3 comments

No ping yet

  1. Chris Southan says:

    But Sean, this was in PubChem from 2005 ? http://pubchem.ncbi.nlm.nih.gov/rest/chemical/173577

    1. sean says:

      Cheers Chris,
      Yes and the Synonym is wrong. Should be ‘number-RP’ and not ‘numberR.P.’ based on the paper I reference.

  2. Chris Southan says:

    This an interesting case, and worth tracking back as to which synonyms moved through the secondary sources with which structures. I can’t square the circles because the CID predates all the SID dates, including the ChemSpider one (can David/Tony tell us the entry date and if it was pulled over and pushed back in 2007 ?) It looks like MeSH annotators have R.P. > -RP and made the (false) Griselimycin linking back to 1981. Not that anyone needs castigating because PubMed authors have used just about all possible syntactic mutations (http://www.ncbi.nlm.nih.gov/pubmed?term=%2227753R.P.%22%5BSupplementary%20Concept%5D)

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>