In the history of developing ChemSpider we have undertaken some fairly demanding curation activities. For example, Vancomycin and Ginkgolide B. Now we are in the middle of trying to resolve the structure of Digitonin. There are 25 (!) skeletons for digitonin on ChemSpider from various sources. There were eleven compounds on ChemSpider called Digitonin. We have been able to clean most of these by removing partial stereochemistry. We are now left with three structures…simply search Digitonin on ChemSpider and you will see three structures with full, but different stereochemistry.

What is a “correct structure” is a matter of assertion. Who says what is correct? What publications, what techniques, what database, who says its correct? Structures have timelines…they can change with time as new analytical techniques are applied.

This is a call to the community to help resolve the existing confusions around Digitonin on ChemSpider…but they are out there in all the other databases also and there are discrepencies between Wikipedia, DSSTox, ChEBI, PubChem and so on. So, my call to community…what is the correct structure of Digitonin and based on what assertions?

With this information in place, and assuming communal agreement on the conclusion, we can go help clean up the other databases. Help!

Stumble it!

29 Responses to “A Request for a Crowdsourced Investigation of Digitonin”

  1. Egon Willighagen says:

    Nice use of the InChIKey!

    I suggest the use the tag csdigitonin for all the crowd sourcing, and that if your start using a new social medium, you add the URL here.

  2. Egon Willighagen says:


  3. Egon Willighagen says:


  4. Egon Willighagen says:


  5. Joerg Kurt Wegner says:


  6. david sharpe says:

    I wonder if part of the problem is that there are 2 (or more) glycoforms that are ‘valid’. I suggest this as I had a look in the Merck index, where it mentions ‘Purification of commercial digitonin and it’s separation into two fractions: G. Ruhenstroth, Z. Physiol. Chem., 1955, 302, 111′ (I’ve not been able to access this article). And the structure shown has one sugar which has appears to have one undefined centre. I was just guessing that perhaps improved separation techniques have meant that what was originally thought to be a single compound has been determined to be 2 or more components and each of them has kept the historical name digitonin.

  7. Kirill Degtyarenko says:

    “There are 25 (!) skeletons for digitonin on ChemSpider from various sources.” brings 25 hits… but these are for vancomycin.
    Please replace it in your post with
    that’s “only” 16 hits in ChemSpider.

  8. Kirill Degtyarenko says:

    Following David’s comment. From “Concise encyclopedia biochemistry and molecular biology” (ISBN 3110145359), 3rd Edition, Walter de Gruyter, 1997, page 174:

    “Digitonin: a mixture of four different steroid Saponins (see) from the seeds of the purple foxglove, Digitalis purpurea. The main component, also called D. [Mr 1229.36, m.p. 235 °C, [α](20)D -54° (c = 0.45 g in 15.8 ml methanol)], makes up 70-80 % of the mixture. The aglycon is digitogenin [(25R)-spirostan-2α,3β,15β-triol, Mr 448.62, m.p. 296 °C (d.), [α](10)D -81 ° (c = 1.4 in CHCl3)].”

  9. ChemSpiderman says:

    Here’s the SINGLE structure for Digitonin on Wikipedia:

    On ChEBI:

    and if you check MeSH its one structure too…

  10. Kirill Degtyarenko says:

    It looks like both ChEBI and Wikipedia structures are wrong as far as aglycon is concerned. According to
    “With our investigations, it was possible for the first time to confirm beyond all doubt the structure suggested by Tschesche and Wulff for digitonin by means of modern NMR techniques, and to assign all proton and carbon resonances.”
    Structure 1 shows methyl group at C-20 going UP, i.e. 20β (while by default spirostan is 20α).

    I was not able to get to the full text of Tschesche and Wulff but an abstract
    “Für Digitonin wurde aufgrund der Strukturbestimmung der erstmalig isolierten Oligosaccharide Digitotetraose und Digitotriose B sowie der Ergebnisse der Methylierung die Konstitution eines 3[β-D-Glucopyranosyl(I)(1→3Galakt.II)-β-D-galaktopyranosyl(II)(1→2Gluc.III)-β-D-xylopyranosyl (1→3Gluc.III)-β-D-glucopyranosyl (III) (1→4Galakt.IV)-β-D-galaktopyranosyl (IV) (1→3-Digitog.)]5α,20βF,25α Spirostan-triols (2α,3β,15β) ermittelt (siehe Abb.1).”

  11. Markus Sitzmann says:

    I redraw the structure from (with a little help by OSRA)

    You can download the SD file from

    The InChIKey of the structure in the SD file is:


    These are the structures for which this Std. InChIKey has been registered in our database (disclaimer: this part of our service is hell of buggy :-) ):

    It includes one of the three structures you suggest, Antony – but I don’t ask me whether this is the correct structure

  12. David Sharpe says:

    I happened to find the following article, which provides a single structure and provides 1H, 13C and various 2D NMR experiments to support its assignment.

    Structure Investigation and Proton and Carbon-13 Assignments of Digitonin and Cholesterol using Multidimensional NMR Techniques
    Petra Muhr, Werner Likussar, Manfred Schubert-Zsilavecz, Magnetic Resonance in Chemistry, 1998, 34, 2, 137

    DOI: 10.1002/(SICI)1097-458X(199602)34:23.0.CO;2-Q

    I hope this is useful.

  13. David Sharpe says:

    I’ve had a quick look at the full text of Tschesche and Wulff, and while I have to admit that the very little German I know, is not enough to pick out anything but the most basic details, I think the Mag. Res. Chem. reference is far more useful as it provides the data on which the structural assignment is based on. Whereas, looking at the Tschesche and Wulff paper it is not clear to me how they derived the assignment that they made.

  14. Antony Williams says:

    Marcus, I agree with the structure you drew for Digitonin as matching the one at So, the question now is whether it’s the correct structure? We have seen some papers referenced on someone willing to do the work to draw the structure from those papers and confirm the one on CommonChemistry as being consistent?

  15. Kirill Degtyarenko says:

    Dear all,

    I did redraw the structure from the scratch looking at Muhr et al. (1998) paper, structure 1.


    It is different from the one in Common Chemistry because, as I said before, the methyl group at C-20 goes up, not down.

  16. Markus Sitzmann says:

    I am currently trying to draw the structure from the Muhr paper which David posted … at least, I managed to got the first InChI block right already … but I know that is not what we are looking for :-)

  17. Markus Sitzmann says:

    Yes, the C-20 group has opposite direction to the commonchemistry structure. However, there is another
    problem with one of the rings – but I might be already gaga enough with this structure to oversee the obvious (or even worse, do a stupid mistake I just don’t see). Here is some explanation:

    Here is an SDF containing the structure I extracted from the Muhr paper:

  18. Chemspider and Structure of Digitonin « Mat Todd's Chemistry Blog says:

    [...] molecule is a “saponin” meaning a steroid with sugars attached. The details are on the Chemspider blog and there is a discussion on [...]

  19. Mat Todd says:

    Finally got round to a) starting my blog and b) posting my thoughts on this issue.



  20. Kirill Degtyarenko says:

    I wonder if anybody can get hold of the following dissertation:
    The structure and stereochemistry of digitogenin. (1956)


  21. Markus Sitzmann says:

    Nice article, Mat!

    However, from your article I saw that the Wikipedia structure lacks a stereo center compared to commonchemistry

  22. Kirill Degtyarenko says:

    I think that the paper “The Constitution and Stereochemistry of Digitogenin” should be a concise version of the dissertation I mentioned in the previous post .

    Digitogenin (structure Ia on page 3169) is drawn with methyl at C-20 going DOWN and hydrogen going UP (convention here is, solid line = UP, dashed line = DOWN). Unexplicably, when I is transformed into structure XI or XIII, the same methyl group is pointing UP.

    I find nothing in the body of the article but the footnote (9) goes like this:
    (9) Gitogenin and tigogenin have been interrelated via gitogenic acid (III) as well as through Δ2-22a,25a,5α-spirosten (VIII) (see ref.
    24). Since tigogenin can be obtained from diosgenin (cf. R. E. Marker, T. Tsukamoto and D L. Turner, THIS JOURNAL, 62, 2525 (1940)), all of the arguments pertaining to the configuration at C-20 and C-22 of diosgenin (for leading references see H. Hirschmann, F. B. Hirschmann and J. W. Corcoran, J. Org. Chem., 20, 572 (1955) , and M. E. Wall, Experientia, 11, 340 (1955) ) also apply to digitogenin. V. H. T. James (Chemistry & Industry, 1388 (1953)) has related the configuration of C-25 in diosgenin with D-glyceraldehyde.

    I am not sure that “leading references” will lead us to any direct evidence of configuration of digitogenin at C-20 but still I think we should check them out.

  23. Markus Sitzmann says:

    Some comment to my comment.gif graphic above: forget the blue part – this was a mistake by me (changing the flat tire at my wife’s minivan in the rain just gave me enough fresh air to finally see my error )

    Besides that, I agree with all the things Mat Todd posted in his blog. The original structure (which I called Muhr structure) seems to be identical to the “Magnetic Resonance in Chemistry 1996″ structure posted by Mat, however – as Mat writes – this structure is not identical to the commonchemistry structure.

    Here is my corrected form of this structure as SDF

    Std.InChIKey of this structure is InCHIKey=UVYVLBIGDKGWPX-DXBRUFHKSA-N

  24. Markus Sitzmann says:

    Interesting, the Wikipedia structure of Digitonin which lacks the definition of a stereo center (as I mentioned above) appeared only recently on Wikipedia (the change appeared Sept. 6; before that date a better structure was shown).

  25. Egon Willighagen says:

    ChemSpiderMan: did this exercise result in a common agreement? What’s the link to the correct ChemSpider entry?

  26. Kirill says:

    Does not look like that :)

  27. Markus Sitzmann says:

    What is the final status of Digitonin?

  28. Antony Williams, ChemSpiderman says:

    Markus….where we have ended up, as far as I can tell based on everything I have followed…is we have a set of extracted assertions/opinions about what the structure might be. I am stuck as to what the final structure of digitonin is actually. I know you’ve followed this discussion quite closely. What’s your opinion? Do you feel we have an answer as a result of all the work that has been done? If not it may well be worth doing the work to figure it out. That would mean some NMR work and I am always up for that :-)

  29. Markus Sitzmann says:

    Well, as I can see, ChemSpider shows now a specific structure for Digitonin. I would say, let’s go with it – however, a cool feature in ChemSpider would be if the Digitonin page would indicate that there was a discussion about this structure (even cooler would be if the entire discussion could be recallable – unfortunately, it went out in so many places).

    And regrettably, I can not help with NMR work … my knowledge about NMR is old an rusty … and that is probably a flattering way to describe it :-)

Leave a Reply