Skip to main content

Infectious Diseases

Selenium In a Drug Structure: Why Not?

You don’t see too many drugs with selenium in them, that’s for sure. It’s one of those elements that can be used to illustrate the Paracelsian doctrine that the dose makes the poison: selenium is an essential element that’s also toxic. There’s no doubt at all about either of those properties; it all depends on how much of it you get.
And that’s the problem with using the element in a drug molecule – the dose of many pharmaceuticals would then exceed the safe amount of selenium that a person could take in. That’s especially true for whopping-dose areas like antibiotics (Home of the Horse Pill reads the sign over the door). So it’s especially interesting to see that Achillion has spent some time and effort developing just that: a new antibiotic candidate whose essential feature is a selenium substitution.
No, they’re not idiots. In fact, I have to salute them for having the nerve to go down this path. The key here is that the selenium in tied up in a heterocycle, a selenophene (analogous to thiophene, and not a heterocycle that very many chemists will have seen.) This keeps the element from being bioavailable, as is apparently the case with the even stranger heterocycle ebselen.
And going from a thiophene to a selenophene is not a neutral switch – in this case, it seems to have been quite helpful. The structures are in a family of topoisomerase/gyrase inhibitors that have shown a lot of promise, but have dropped out of development due to potential cardiac side effects. It’s the dreaded hERG channel again, which has sunk many a development program. Binding to that ion channel can lead to long QT syndrome in some patients, and you really don’t want that risk. (Neither do the regulatory agencies, which require testing of any new drug candidate for just this reason).
Switching to selenophene gave the cleanest hERG profile for Achillion’s entire series of compounds, while still retaining antibacterial activity. So these selenium heterocycles are, for the adventurous, probably worth a look – they can be similar to thiophene in some situations, and not so similar in others. People are going to look at you funny if you make them, but you should never let that slow you down.

30 comments on “Selenium In a Drug Structure: Why Not?”

  1. tuky tuky says:

    Is there a link for this publication?

  2. CMCguy says:

    What’s next Arsenic based drugs returning? Paul Ehrlich would be proud.
    Definitely interesting and worth the salute. I wonder about the stability due to oxidation potential however am guessing the heterocyclic ring modulates that.

  3. Rick says:

    As most of us know, selenomethionine has been a godsend for crystallographers. Bacteria eat it up and make proteins with nice heavy atom derivatives at defined positions without seeming to suffer much.
    In a world where boron, deuterium and even arsenic (for you old-timers) have been incorporated into drugs with promising results, it’s kind of hypocritical to look at someone funny for substituting covalent Se for covalent S. Hats off to Achillion!

  4. luysii says:

    The long QT syndrome is nasty indeed, and not something you want to precipitate with a drug. Just about all cases not due to drugs are hereditary and due to ion channel mutations (over 10 are known). For just how nasty long QT can be see —

  5. okemist says:

    As for old school chemotherapeutics: the organic arsnic compound Darniaparsin is just finishing PII trials for hemological cancers.

  6. Nick K says:

    The rest of the molecule is rather curious as well. Is it derived from quinine/cinchonine?

  7. Ron says:

    #7: Yes, the paper Derek linked to indicates that the compound is synthesized through degradation of quinine and then derivatized.

  8. Membranes R' Us says:

    Look in JMed Chem. March 28, 2011. Cool molecule, data, not so. SAR looks flat in all sensitive and resistant strains and acts like a typical membrane perturber. AND it doesnt touch Gram negative bugs or the mutant resitant FQ strains.
    Pass this compound through a membrane depolarization assay or better yet, macromolecular synthesis studies to really impress us!
    But I think J Med Chem is where once hopeful chemicals go to die peacefully- Our own molecular graveyard!

  9. partial agonist says:

    but it is not a Canadian singer

  10. Derek Lowe says:

    Partial, not even Angewandte Chemie would touch that pun. At least, I hope no one ever sends them a paper about that structure so that they’re not tempted. Yeeesh.

  11. JasonP says:

    Other than selenium, what other non-standard but likely safe elements can you put into a compound? Something as indelible as an electron cloud could make the difference between signal bias etc no?

  12. MoMo says:

    What’s this?
    Idi Amine
    Any silicon based compounds in the PDR put there intentionally?
    Good post Memebranes R’Us- couldn’t have said it better.

  13. jgault says:

    I see a new biotech on the horizon
    “Selenicon Therapeutics”.
    Hell, Velcade sells what $1.5 Bil/year, why not Selenium, just make the stuff in a lab a long way away from me.
    Damn the dogmatic, full speed ahead.

  14. susurrus says:

    This is going to absolutely mess with my Pipeline Pilot protocols 😉

  15. Mouse says:

    Looking at the paper, the hERG data is unconvincing in this class the Se does anything useful as other more normal analogs have similar hERG. We should be careful not to accept as dogma the assertion that selenocycles will fix hERG.

  16. Libby says:

    I wonder if the drug would stink? A lot of selenium compounds smell pretty bad.

  17. NoDrugsNoJobs says:

    Interesting post Derek – might be an interesting substitution for IP purposes in some instances…

  18. drug_hunter says:

    From an NIH web site:
    “Brazil nuts may contain as much as 544 micrograms of selenium per ounce … It is wise to eat Brazil nuts only occasionally because of their unusually high intake of selenium.”

  19. lynn says:

    #8 Membranes R’ Us – While I agree that it would be good to see membrane depolarization, the effect of serum on MICs (to compare with cytotox data, done in the presence of serum) and macromolecular synthesis, the Se compounds are quite active against the multiply FQ-resistant strains (when compared to FQs). You might say the SAR is flat across the FQ (sensitive and) resistant strains, but that’s as would be hoped for Gyr/Grl inhibitors that are not cross resistant with FQs. I don’t get a membrane-active vibe from these data. But, as noted by Mouse, hERG isn’t very impressive

  20. grumble says:

    I wonder if the tellurophene analog has any activity? Pure tellurophene supposedly possesses a terrible stench, thought it is air-stable!

  21. petros says:

    Surely the structure to aim for is a silicon-lithium complex perhaps with some Vanadium thrown in for good measure!
    SiLi or V SiLi

  22. David Formerly Known as a Chemist says:

    hERG channel blockers don’t necessarily lead to Long QT syndrome. Tolterodine, a muscarinic antagonist used for overactive bladder, is a 50 nM hERG inhibitor that shows no clinical prolongation of QT.

  23. Marge says:

    @ Libby: Yes, Se compounds do stink. One of my labmates married a Se chemist, and she ALWAYS made him shower first thing when he got home. “If you think sulfur smelled bad…”

  24. Lu says:

    I can only imagine how some of its intermediates smell…
    No, I would rather not to!

  25. drug_hunter says:

    What are some other unusual elements that we should all be thinking harder about? For example Si shows up occasionally. What about phosphorus (other than in phosphate groups)? Other than velcade, do we have any favorite Boron-containing compounds? Etc.

  26. dave b says:

    More tellurium based drugs required! Minor side effects can be mitigated with a clothes peg on the nose.

  27. JL says:

    #25: how about iodo, maybe not as unusual as some, but still pretty rare in final target molecule.

  28. Anonymous says:

    An ortho-thiotellurophene…argggh. What were they thinking. Sell that turd to GSK and reap a big profit. They will polish it and make it shiny and whole…

  29. Anonymous says:

    Given the -COOH moiety in the molecule, I can’t see what Se vs. S would do to improve hERG. Looks like BS to me

  30. Mark Z says:

    I went to a seminar held by Dr. Kip Guy, a researcher at St. Jude’s Children’s Hospital. The seminar was in 2009, and was on approachs to increasing moleuclar diversity. During the question and answer session, I asked Dr. Guy if he had considered selenium heterocycles as a possibilty to increase diversity. His response was that selenium was so very toxic that he didn’t think it would be a useful path to follow. I/m very glad to hear that soneone is investigating these kinds of compounds.

Comments are closed.