Skip to main content

Academia (vs. Industry)

What Does One Do With These?

A reader sends pointed me to this paper, whose title they found interesting: “Characterization of a Steroid Receptor Coactivator Small Molecule Stimulator that Overstimulates Cancer Cells and Leads to Cell Stress and Death”. An unusual mechanism, for sure – so what is this compound, exactly?


You have to go into the Supporting Information to find that out. “MCB-613” was identified in a cellular screen with compounds from a Molecular Libraries Probe Production Centers Network collection. And when you track down the structure, well, there you have it: yet another bis-chalcone. Whoever produced a library of those things really ripped off the funding agencies, that’s for sure.

What’s a bit annoying and concerning is that the paper just takes the structure and runs with it. There’s no mention of the fact that it might not be very drug-like, and that it might be reactive. The paper has a lot of painstaking work in various cell assays to try to work out the effects of the compound, and includes some SPR assays that seem to show reversible binding to one component of the nuclear coactivator SRC-3. But there are a lot of things that a structure like this can do, and there seems to be no recognition of this. It’s a compound, and it hits, and what else do you want? They mention that it’s a chalcone, but as a feature, because lots of chalcones are bioactive, and so on.

There’s similar stuff in this recent J. Med. Chem. paper. The authors have been working in this area for quite a while now, and they’ve not only got the bis-chalcone, but both aryls are Mannich products as well. This is a festival of problematic functional groups, to be honest. And the paper has plenty of cellular activity data, with (as in the other paper) some apparent selectivity between tumor cell lines and more normal ones. But still.


These structures have a very low probability of ever being drugs – I suppose that’s what I’m trying to get across. Now, there are a lot of unlikely chemotherapy drugs out there (nitrogen mustards being one vicious example), and structures like this could possibly still find a use. But there are a lot of assays that would need to be done in order to shore things up a bit. For example, doing a whole-proteome screen to look for covalently modified proteins would be a good idea, since you’d really expect something like that to be going on.

And that’s not even getting to the big question: in vivo. It’s one thing to show effects against tumor cell lines in culture, but the real challenge for these kinds of structures will come when they have to be dosed in an animal. The J. Med. Chem. paper talks about these as “lead preclinical drug candidates”, but nothing’s really a drug candidate until it’s been in animals, not to put too fine a point on it. I know that’s a completely different realm than the authors of these two papers are working in, but – how to put this? – the numbers of compound that kill tumor cells are almost beyond counting. It’s not that high a bar to clear, and if you’ve got something interesting, it needs to be taken through those next steps, rather than (say) have another thirty-four cell assays run on it. An awful lot of work has gone into this sort of chalcone chemistry over the years, partly because you can make a lot of compounds very quickly. But what’s come out of it?

36 comments on “What Does One Do With These?”

  1. milkshake says:

    If you look at the structure of Sutent, you see a condensation product of oxindole with pyrrole aldehyde. This class of compounds actually behaves as very poor Michael acceptors (for example the C=C ignores reactive amines like pyrrolidine) probably due to pyrrole and oxindole rings being very electron rich. Still, developing the “chalcone-like” class of molecule has been painful – very yellow, photolabile (E/Z izomerization on sunlight), often poorly soluble. You don’t see any other Sutent like molecules that made it through development.

    The structure above is probably going to be a thiol-reactive Michael acceptor (which might be OK if it shows any specificity, I would worry abut nonselective reaction with protein SH groups). The compound below is likely to be a promiscuous protein binder, like many polyphenols, and I would worry about quinomethide formation.

    1. Hap says:

      1) Should the aldol adducts above (ketone-aryl aldehyde adducts) be more electron-deficient and more likely to undergo Michael additions than oxindole-aryl aldehyde aldol adducts (since the oxindole is basically an amide/lactam)?

      2) The second compound really looks like it’s going to give birth to twin bouncing quinone methides pretty quickly, and I wouldn’t figure they’d be selective for much once formed (other than nucleophiles, olefins,…) At least its enones aren’t so electron-deficient though (small favor).

  2. d says:

    “Whoever produced a library of those things really ripped off the funding agencies, ”
    sorry, non specialist here, could someone explain what derek menat by this?

    1. Mike says:

      He means someone was paid to make really inexpensive chemical compounds that are unlikely to be very useful for their intended purpose.

      Personally, I don’t see any reason to characterize it as “ripping off the funding agencies”, but I would say that any funding agency that paid for these chemical compounds probably made a poor choice.

      1. b says:

        I believe he means that there is a bunch of research money being used (wasted) by labs like these to pursue compounds that have little to no probability of ever being useful, not necessarily that somebody was paid to make the compounds themselves.

    2. Chembot says:

      Molecular Libraries Probe Production Centers Network was an NIH-funded initiative.

  3. M says:

    Some good news:

    Looks like there is improvement among the surviving volunteers. Two have been sent to other hospitals closer to home, two are still in the Hospital in Rennes but have improved, and the fifth who never showed symptoms has been discharged.

  4. anonymous says:

    Here is another recent paper claiming a specific activity from a natural product with a chemical structure that suggests non-specific interactions (polyphenol, bis-aromatic aldehydes, dna intercalator-like structure.,,)
    Lan Lan, Carl Appelman, Amber R. Smith, Jia Yu, Sarah Larsen, Rebecca T. Marquez, Hao Liu, Xiaoqing Wu, Philip Gao, Anuradha Roy, Asokan Anbanandam, Ragul Gowthaman, John Karanicolas, Roberto N. De Guzman, Steven Rogers, Jeffrey Aubé, Min Ji, Robert S. Cohen, Kristi L. Neufeld, Liang Xu. Natural product (−)-gossypol inhibits colon cancer cell growth by targeting RNA-binding protein Musashi-1. Molecular Oncology, 2015; 9 (7): 1406 DOI: 10.1016/j.molonc.2015.03.014
    Here is the structure of this natural product:

    1. JAB says:

      Well, geez, gossypol! What target is it not active at?

  5. Rule (of 5) Breaker says:

    Wait, a bunch of biologists testing compounds is not the same thing as drug discovery? Who knew? You know, besides all true drug researchers everywhere. I was not at all shocked to discover that the papers were written by academics. Very few academic institutions (notice I did not say no institutions – there are a few) have an understanding of what drug discovery entails. Perhaps they are hoping to license these things, but again this just puts their ignorance on display, as nobody (reputable) in biopharma is touching these molecules.

    1. Anon says:

      Pardon me, but that sounds quite arrogant. At least when academics push forward non-drug like molecules they fail fast and cheap in preclinical, whereas Pharma’s drugs fail later, in Phase 2/3 or even in the market after spending hundreds of millions of dollars. The result is the same: zero.

      1. Anon 2 says:

        Not to mention the fact that SMEs and academia are the source of most new drugs:

        1. Rule (of 5) Breaker says:

          These are exactly the responses I would expect from academics. Academia the source of most new drugs? That is a nice article, other than it is completely missing any references whatsoever, and the methodology used to determine where the drug came from is not communicated. Arrogance would have been claiming that all academic institutions are incapable of drug discovery. Note, that I said most, which I challenge any industry drug reseacher to dispute. I am guessing I will get a lot of agreement on that one. As far as failing fast, that is a given with garbage compounds like these.

          1. Anon says:

            So how come Pharma is desperately pursuing open/external innovation initiatives, partnerships and pipeline agreements with academia, even relocating their head offices and opening research centers on academia’s doorstep … you know, having admitted that its own internal R&D can’t deliver crap?

          2. b says:

            Please don’t delude yourself into thinking that the appearance that management knows what they’re doing is the same as them actually knowing what they’re doing. They’re doing it because it’s the new cool thing to do and everybody else is. And because using grad students/postdocs to do all the work is way cheaper than actually having to pay a full employee. Almost all of those making these decisions will be long gone before it’s known whether it works or not. Anybody that’s been in the industry long enough knows it’s an endless string of cyclical fads that management flocks to and sells as the amazing new way to drastically improve output. In the end it’s all more smoke and mirrors.

  6. The xkcd reference that’s relevant here: .

    1. Pennpenn says:

      I take it you mean page 1217 (, called “Cells”. The quote being “When you see a claim that a common drug or vitamin “kills cancer cells in a petri dish”, keep in mind: So does a handgun.”

  7. Just Asking.... says:

    I don’t have a subscription, so I can’t read the original article, but I read the summary. My question: are they trying to suggest that MCB-613 is a potential lead (which I doubt, if it’s only identified in the backup material), or are they trying to suggest that overstimulating SRCs will negatively select against cancer cells? Is that so obvious that it wasn’t worth publishing?

    Is it a proposal for a mechanism to attack some cancers that another research team could advance? Or are they just announcing that a hammer can bang in a nail?

    1. anonymous says:

      I don’t have access to the paper either, but I think everyone is missing the point. The interest thing about this paper is that SRC3 is druggable and potentially disease modifying, not that MCB-613 itself is a drug lead. MCB-613 is only a tool compound. The only question is it an adequate tool compound (i.e., not a false positive).

      1. Bagnar says:

        This paper is accessible through a “famous” russian website you can find by clicking on my name above.
        Have a good reading.

  8. Anon electrochemist says:

    Derek, I hear reviewer suggestions like yours often, but never how should they be implemented. Every “whole proteome assay” I’ve ever seen has been a two week fishing expedition with valuable Orbitrap time. Any advice from fellow readers much appreciated.

    1. alchemist says:

      I guess the only suggestion is not to make excessive claims when the tool compound has such high risk of giving spurious results. Unless you are indeed prepared to spend a few weeks with your Orbitrap. If the only tool compounds coming off a scren are of this type, you should give a lot of thought about the reliability of the data you will generate, and whether it is really woirht generating it with this tool.

  9. Hap says:

    Anon 7:20 – because they want cheap, and student labor and startups (who aren’t paying the wages or benefits that they would have to) can give them that. Productivity doesn’t appear to factor into it (else why invest in a set of companies that have generally been an investment sink) – getting quick money and getting out are all that matter.

    1. Anon says:

      Well that sounds like a very sensible business decision compared to pouring billions into internal R&D with little to show for it.

    2. Anon says:

      PS. “Productivity doesn’t appear to factor into it (else why invest in a set of companies that have generally been an investment sink) – getting quick money and getting out are all that matter.”

      I think you just contradicted yourself, given that productivity *IS* what you get out vs what you put in, no?

      1. Hap says:

        1) Then why play there? You can’t make drugs without spending lots of money on research. You know what your research can do – why do you think the stuff you don’t know is going to work better? Based on the flood of papers like this (with not so much awareness of what kinds of structures might be problematic), and the flood of money lost in biotech, I don’t know why you’d think you’d do better. If you don’t know what your research can do, of course, you also probably shouldn’t be playing there – there’s a reason you don’t let everyone play with high voltage.

        2) I mean that the people running things don’t necessarily care what the research puts out – the stock bump is what they want, and what happens to the research productivity will not be their problem.

        1. Anon says:

          “why do you think the stuff you don’t know is going to work better? ”

          Nobody does, but exploration is a prerequisite for discovery. And doing what we’ve always done certainly won’t work any better than what we’ve always done.

          At least academics are prepared to challenge the status quo, and that gives them an edge to discover something new.

          1. Hap says:

            If you don’t know what you’re doing, why should anyone else think that what you’re doing is going to be worthwhile? The expectation for decent research is that you have an idea what’s been done before and, if your research would be expected to be questionable on that basis, to explain why the previous work might be wrong or inapplicable. Doing crazy things without any awareness that they’re crazy suggests either that you don’t know what’s been done before or you don’t care, and in either case, I don’t see why anyone would assume that the products of said research would be at all worthwhile, let alone more likely to suggest drugs, than previously known work.

  10. entropyGain says:

    This is a common complaint amongst the true drug hunter community, and I’ve seen a lot of it come out on this blog. In part, it is our fault for not actively participating in the peer review processes. I’ve sat on many NIH study sections and am usually alone or one of a small minority of scientists on the panel with industry experience.

    If real drug hunters took the time to sit on more study sections or review for journals, then the community as a whole could be steared to more viable avenues of translational research. We may think there is not enough time or not enough payoff. However, this is a great way to learn the emerging state of research and exercise our neural networks outside our normally narrow areas of specialization. And we all waste plenty of time in non-productive activities that if pruned could provide enough time to review a few grants or papers.

    As a side note, the paper under discussion here is in a quality journal from a first tier lab in the field, so it’s not just an aberration…

  11. MoMo says:

    “And we all waste plenty of time in non-productive activities ”

    Like attack a bunch of student Texans over a bis-chalcone . O’Malley would eat each of you “drug hunters” for breakfast.

    Now get back to work making amide bonds and pondering the latest fad science CRISPR

  12. hn says:

    It’s not about drug hunters or industry experience. It’s about biologists ignoring the importance of chemistry.

  13. FooBar says:

    Garbage. Bert O’Malley or not, Garbage.

    Just compare Fig. 1A — “dose-response” of the bis-chalcone stimulating SRCs) — with Fig. 1G — binding of bis-chalcone by SPR.

    Fig. 1A: minor effect at 5 uM, large effect at 6 uM, huge effect at 7 uM, and enormous effect at 8 uM. Quiz Question: What’s the Hill coefficient for these effects?
    Fig. 1G: Essentially no binding until ~15 uM, with half-maximal binding ~??? uM — Oh, it doesn’t ever plateau.


    1. Laughing says:

      Thanks for drawing attention to that figure and the ludicrous concentrations they chose to test the compound at. Likewise Fig 1C-F (4, 6, 8 and 10 uM), Fig 2A (in cells this time) 3, 4, 5, 6 and 7 uM.

      I’d be laughing my arse off, except that based on that data they went and did a xenograft experiment on 27 mice. Now, I’ve nothing against careful animal experiments in the cause of good science, but am I alone in thinking this is the kind of thing that gives science a bad name?

    2. tangent says:

      Fill it in for a layperson if you would, what type of non-specific mechanic does that data say to you? Aside from the whole fact that we’re talking micromolar at all.

      Fig. 1A, so I looked up the Hill equation that curve looks like positively cooperative binding, okay, but then what are you concluding out of that?

      Fig. 1G… says that whole Fig. 1A effect ramp-up was at levels before it was even binding? That wouldn’t make any sense.

      The no maximum binding means what, the compound is off hitting on other things indiscriminately? But isn’t this binding assay run against the purified (putative) target, so what else is the compound getting off to, stuff like the experimental apparatus?

  14. MoMo says:

    What gives science a bad name is negligent reviewers and publishers that fill their coffers with articles that bring into question of the value of the journals themselves.

  15. Mandrake says:

    Q. What Does One Do With These?
    A. Use them as model rocket fuel!

Comments are closed.