Skip to Content

Alzheimer's Disease

In Which I Hate A Wonder Drug

Schering-Plough has had its share of troubles over the years, but the company has also seen itself saved by some pretty unlikely compounds. Vytorin (ezetimibe) is the example I’ve spoken about here, and if the drug doesn’t seem like a savior at the moment, well, you have to keep in mind that it was the biggest thing for them since Claritin went off-patent ten years ago.
Now there’s another one potentially coming up. Expectations are building for a thrombin receptor antagonist compound, SCH 530348. And I have a history with this one, too: while the labs down one hallway from me were discovering ezetimibe, down the other hallway they were laying the foundation for this one. There’s a big difference, though, in the way I saw the two.
This thrombin antagonist is an unlikely drug for several reasons. For one thing, its structure is not the sort of thing most medicinal chemists would go out of their way to make. But there’s a good reason for that: to a first approximation, it wasn’t made with medicinal chemistry in mind. 530348 is based on a natural product called himbacine, whose fame, such as it is, rests on its properties as a semi-selective muscarinic antagonist. And that’s how Schering-Plough got interested in this class of compounds; thrombin had nothing to do with it.
At the time (early to mid 1990s) the company had a team working on Alzheimer’s disease, and I’ll go ahead and mention again that I was one of the people involved. (Five minutes on SciFinder would tell you that, anyway). We were quite interested in selective muscarinic antagonists, particularly for the m2 subtype, and himbacine was at the time one of the more selective compounds with that profile. So one of the group leaders at the company, Sam Chackalamannil, decided to synthesize it and do some SAR around the structure.
That was no small undertaking. Himbacine’s not one of the most complex natural products by any means, but it’s no stroll to the beach, either, especially when compared to the usual sorts of drug structures. It took a lot of time, a lot of ingenuity, and (most importantly) a lot of effort to do it. And I. . .well, I thought this was a terrible idea.
I really did. By the time himbacine itself got made, the project team had muscarinic compounds that were more selective and more potent (and a lot easier to make, to boot). I would listen to Chackalamannil’s people presenting their long, difficult routes during meetings, and I’d sit there imagining the company going slowly bankrupt if everyone adopted this approach, the revenue slowly sinking as the number of JACS communications rose. I couldn’t see the point, and although I don’t think I ever quite had the nerve to say so to Chackalamannil himself (hi, Sam!), I said it to plenty of other people.
So, is it time for me to eat crow? Well, one plateful, at least. Some of the himbacine analogs hit in the high-throughput screen for thrombin activity, to everyone’s surprise, and some further compounds (now shed of their muscarinic activity) were even better. The drug discovery effort culminated in 530548, which now might be about to benefit a huge number of people and make the company a ton of money, if everything goes well.
Of course, if these things hadn’t hit in the thrombin assay, I could have remained secure in my opinion. After all, they were never worth very much as muscarinics, as far as I know. (Of course, our muscarinic compounds, in the end, never were worth very much as Alzheimer’s drugs, which is something to keep in mind). So that’s the question: how likely is it for molecules like this to work? It’s very hard to answer that, but given this data point, I guess the answer is “at least a little more likely than I thought”. The very fact that they didn’t look like most other things in the screening deck was probably in their favor. I still think that these compounds were a long shot, but this is a business that lives on long shots. This one came through, and congratulations to everyone involved.

8 comments on “In Which I Hate A Wonder Drug”

  1. Nick K says:

    Hats off to the chemists who made Himbacine, and to a very brave management. Once again, this saga illustrates the importance of serendipity in drug discovery.

  2. Jack Friday says:

    How convenient – just when the wheels fall of the Vytorin waggon another hopeful appears over the horizon.
    Call me an old sceptic……..

  3. Derek Lowe says:

    OK, JF, you’re an old sceptic.
    Actually, this compound has been in the works for years. These showed potential thrombin activity back around the time I left Schering-Plough, which has been years ago. That would make it about six or seven years past the discovery of ezetimibe, and allowing for that compound’s difficult birth, that’s just about right on target.

  4. milkshake says:

    I just looked up their JACS paper on total synthesis of (+)-Himbacine and it is very nice.
    I think Himbacine is reasonable as a natural compound-derived scaffold, it does not have any funky groups and the Diels-Alder approach fairly modular. The only complaint is that the molecule looks somewhat greasy. But many statins and steroids are funkier. (Acetylene-containing contraceptives for example!) I agree that cranking out biaryls by Suzuki reaction is faster – yet sometimes you get lucky with not-so-simple compounds.

  5. PorkPieHat says:

    As drug companies divest of natural products, the industry loses opportunities from “the wisdom of the ages” that is the serindipity to which Nick K refers. If you and I ever meet, D.Lowe, you’ll have to ask me about another “Himbacine analog” – like story that is a billion$ drug today. If SOME fraction of medicinal chemists are not permitted the freedom to do what Chackalamannil did within our industry, we don’t benefit from this high-risk high reward activity.

  6. Panoramix says:

    The molecule looks like a hybrid, the southern part is totally artificial and the northern part is pure total synthesis of natural products.

  7. RTW says:

    As a 20 year medicinal chemist I have worked on unlikely structures like this myself. In Anticancer drug design much was derived from difficult to prepare natural products. VP-16, Tenopiside, Campothecin, Mitomycin-C, its not unprecedented at least in Anti-bacterials, Anti-Cancer Agents and Anti virals, though even in these areas more “traditional” heterocycles are taking their place…
    I have to agree also that these seem very unlikely in CNS, and CV therapies. But one must remember that Mevacor started out as complex natural product as well.
    At any rate – For the actual chemists in the group. I found a C&EN article that contains a nice blurb and the structure.
    http://pubs.acs.org/email/cen/html/021806150931.html
    Enjoy.

  8. vm ganata says:

    Well, in med school pharmacology, they love to repeat the story of Fleming’s serendipitous discovery of penicillin. I’m not surprised not much has changed in terms of drug discovery.

Comments are closed.