1. petros on June 1, 2012 7:46 AM writes...

I blame the journal's editorial board memebrs! :)

2. Mike Pollastri on June 1, 2012 7:58 AM writes...

I think that there should be some mechanism by which known promiscuous chemotypes, toxicohopres, etc should be flagged upon submission for publication. One way to do this is to require that any compounds that reported with in vitro screening data be submitted to journals be submitted as an auxiliary file in SMILES format so that the journal could (automatically, upon electronic submission of the article) generate a report for the reviewers. Further, since Scifinder is so good at generating calculated properties when you do structure searchers, how about including these PAINS alerts, along with some other flags?

I think education about these kinds of things is most important. Making it so it hits people in the face when they are reviewing papers like this, or when they are doing literature work on a chemotype of interest, would go a long way, I think.

3. anon on June 1, 2012 8:12 AM writes...

Esteemed community:

What key criteria/experiments should a rhodanine- containing compound meet before you develop it further as a tool/probe compound?

Does anyone know of any useful, validated rhodanine-containing probe compounds ?

4. Virgil on June 1, 2012 8:34 AM writes...

Regarding the dynamin effects reported in the ACS MCL paper, color me unimpressed. An IC50 of 7 uM is nothing to write home about for most cell biologists.

5. John Wayne on June 1, 2012 8:38 AM writes...

If I was reviewing a paper that described a rhodanine-containing tool compound, I would be looking for a clean off-target toxicity profile appropriate to the biological activity of interest.

6. Rick Wobbe on June 1, 2012 8:44 AM writes...

Depending on the "scientific literature" you're talking about, the sweeping statement that these compounds are "polluting the scientific literature" seems narrow-minded. I agree that certain compound classes (the rhodanines among them) warrant substantial skepticism with respect to their medicinal or therapeutic potential, which should influence whether and how they appear in medicinal chemistry literature. But broadening that to the entirety of scientific literature ignores the importance of toxicology, which I hope you don't intend to do. Arguably, we have learned more about metabolic and physiologic processes from studying toxins than from drugs, the difference between the two being one of degree, not kind. Furthermore, the design, synthesis and testing procedures we reflexively associate with "medicinal" chemistry are equally apt for toxins; it just depends on the endpoint you're vectoring toward. As such, their value as so-called "tool compounds" should not be belittled. And I am saying that with a straight face.

7. partial agonist on June 1, 2012 9:12 AM writes...

People arguing that these compounds & reports are OK because the authors are making "tools/probes, not drugs" miss the fundamental point that a molecular probe needs to be very selective, first of all, if anyone is to learn anything of significance about the perhaps unvalidated target in question.

Given their well-known polypharmacology, the natural assumption to make about any rhodanine reported is that it is inherently non-selective, unless the authors very clearly prove otherwise.

If they don't prove otherwise, then "polluting the literature" may be a little strong to say, but the authors are certainly ignoring certain fundamentals of medicinal chemistry familiar to all of us, in industry (where I used to be) or academia (where I am now), and reviewers should call them on it.

8. lynn on June 1, 2012 9:44 AM writes...

I’d say it all depends on what the “tools” are to be used for. If you are just looking at the interaction of an inhibitor/binder and a protein/receptor, then selectivity is not critical. But once you’re working in a biological system where your tool can add invisible screws and wrenches into the mix, then it loses its value. In target-based discovery, there should be no assumption that any physiological effect of your compound is due to its interaction with your target of choice. This is particularly true in the antibacterial area, where screening large chemical libraries for inhibitors of specific enzymes may yield hits, but any antibacterial activity of these hits is seldom due [solely if at all] to inhibition of the screened target. As Rick W notes – tox [off-target activity] and desirable target-based activity are a continuum; at higher concentrations, selectivity may well be lost. But it is up to the investigator, academic or industrial, to substantiate selectivity within the relevant concentration range.

9. Pete on June 1, 2012 10:15 AM writes...

Anyone done cyclic voltametry on rhodanines?

10. Anonymous on June 1, 2012 12:06 PM writes...

At least this compound is real.
Just look on this graphical abstract.
I think authors just made it up.
http://pubs.acs.org/doi/abs/10.1021/jm201442t

11. Practical Fragments on June 1, 2012 12:59 PM writes...

Practical Fragments did an April Fools' post on rhodanines recently:

http://practicalfragments.blogspot.com/2012/04/universal-fragments.html

Sadly it looks like too many people still don't get the joke....

12. Hap on June 1, 2012 1:54 PM writes...

People want tools as compounds with known activities and specificities from which they can determine the activities and selectivities of other compounds. If it can't do that well, then either it won't be useful or it will lead people to think they know what they don't (or to know something that isn't so).

If something is a member of a class of compounds known to be not very selective for a specific class of enzymes, then you can at least test the compound against members of the class to see if there is any selectivity (if that wasn't how you found it in the first place). Even then, the compound may still be questioned because other unknown enzymes could be lurking and binding your compound, but you may at least know what to look for. On the other hand, rhodanines seem to bind lots of proteins (and lots of types of proteins), so where do you begin or end in finding what potential targets they don't bind to?

13. Ted on June 1, 2012 2:21 PM writes...

Ho ho ho...

These are funny compounds!

Any industrial program/project leader with so dysfunctional a screening hit BS detector, leading a group of chemists that are similarly ill-equipped (or unable to rebel) deserves the likely fruit of these labors.

Industrially, this frequently occurs at the intersection of inexperience and hubris.

In fact, compounds of this sort are probably good probes to use in a diagnostic assay for overconfident, inexperienced experimentalists...

-t

14. partial agonist on June 1, 2012 4:51 PM writes...

#10

At least the authors did say that the rhodanine compound shown has "features suggestive of pan assay interference compounds, which can correlate with undesirable chemical properties" and they gave some selectivity data.

To my eye their isoxazolone compound (5) looks at least a little more palatable than the rhodanine.

15. dora the explorer on June 2, 2012 8:38 AM writes...

#3: ponesimod (actelion's S1P1 agonist) is rhodanine-like, and is currently in Phase II clinical trials:

http://jpet.aspetjournals.org/content/337/2/547.full

16. Psychophant on June 2, 2012 1:08 PM writes...

I disagree with #6 Rick, and agree with #13 Ted. The term "pollution" is fully justified. Any skepticism due is nothing to do with medicinal potential or the value of "toxins", which is a red herring for this topic. This paper is all-too-typical in not supporting supposed target binding with sufficient rigor. As others have pointed out, rhodanines (and other compound classes) appear in hundreds of papers of this type all purporting to show them as leads against different mechanisms. Is it likely that these rhodanines specifically recognize such diverse targets? No, the MOA is most likely not 1:1 reversible binding. Of course other MOAs might be acceptable, but it should certainly raise eyebrows as to the interpretation of the assay results. Regardless, I doubt that any of the papers on rhodanine activities even bothered looking. ("Why would I need to check my compounds aren't artefacts? I'm sure they're OK.")

Yes, this is pollution, and the cumulative weight of this kind of publication leads to at least three problems. First, and most important, others may wind up using inappropriate compounds as tools, wasting their valuable time. Second, the drive to publish hits without rigorous confirmation fits snugly within an academic culture of "never mind the quality, count the publications". (As evidence, I'd ask what proportion of hit discovery papers go anywhere? How many are followed by lead optimization papers? Relatively few, I'd guess.) Third, it fosters the impression among people such as my employers that there are hundreds of academics who know how to do hit ID. Frankly this devalues those in all types of organization who genuinely do.

Now, I'm not saying that no academic groups know how to do hit ID, but when I was a post-doc at a certain US university we were all sure we did. When I joined industry I was exposed to lots more data & saw how often it all falls through, and now I realize how naive we all used to be.

17. sgcox on June 2, 2012 1:57 PM writes...

#14 partial agonist:
Please look again at the compound (1) featured in the abstruct.
It is not a rhodanine, it is a pure fiction.
It was never made by a human being yet.
Authors claimed it came from Maybridge. Well, Maybridge does have a related rhodamine but not a featured molecule.
Worst of all authors claim the compound was resynthesised. Good luck finding any chemistry in that paper.
Shame on Scripps and shame on JMC and reviewers.

18. Maks on June 3, 2012 3:40 PM writes...

As somebody who left synthetic chemistry for cell biology/molecular biology I consider such articles quite problematic. Most biologists lack the chemical knowledge to judge compounds like the ones above, just buy them from whatever supplier and draw conclusion (and articles) from their effects. A drug for humans doesn't have to be specific but simply has to work, a chemical tool has to be specific to be useful. Please guys, you can do better than that.