Back in 2015, Practical
Fragments highlighted a new resource calling itself “The Chemical Probes Portal.” At the time it included just seven probes, and my post concluded, “I
hope this takes off. Understanding the natural world is hard enough even with
well-behaved reagents and carefully controlled experiments.”
Well, take off it has, as
illustrated by a new (open access) paper in Nucleic Acids Res. by
Susanne Müller (Goethe University Frankfurt), Bissan Al-Lazikani (MD Anderson
Cancer Center), Paul Workman (Institute of Cancer Research), and collaborators.
The paper notes that “the
widespread use of small molecule compounds that are claimed as chemical probes
but are lacking sufficient quality, especially being inadequately selective for
the desired target or even broadly promiscuous in behavior, has resulted in many
erroneous conclusions in the biomedical literature.” As an antidote, the Portal
is an “expert review-based public resource to empower chemical probe
assessment, selection, and use.”
Any scientist can suggest a potential
probe, and these are then internally reviewed and curated. Assuming enough public
information is available about the molecule, probes are then sent to three members
of a Scientific Expert Review Panel for further vetting. Reviewers rate probes
from one to four stars for use in cellular and/or animal models and recommend relevant
concentration ranges. Importantly, reviewers can also include comments to
highlight off-targets, lack of certain data, oral bioavailability, or anything
else.
From a mere seven probes in 2015
the Portal has grown to include more than 500 molecules covering more than 400
protein targets in about 100 protein families. About two thirds of the probes
have three or more stars, meaning they are recommended. The Portal is very easy
to use and can be searched by probe or protein. Laudably, all the data can also
be easily downloaded in bulk.
In addition to the chemical
probes, the Portal also contains around 250 “Historical Compounds” that have
been described in the literature but “are not recommended to be used to study
the function of specific proteins as they are seriously flawed.” These include
molecules such as gossypol, a known aggregator that has been reported as an
inhibitor of multiple proteins, and curcumin. If you see a molecule used as a
probe in the literature, it’s worth checking to see whether it shows up in the
Portal.
The Chemical Probes Portal features
heavily in a Conversation between Cheryl Arrowsmith (Structural Genomics
Consortium) and Paul Workman published (open access) last year in Nat. Commun.
The researchers concisely define chemical probes as “small-molecule modulators
to interrogate the functions of their target proteins, as opposed to protein location,
or other physical properties.” Importantly, they differentiate chemical probes
from drugs. “Drugs don’t necessarily need to be as selective as high-quality
chemical probes. They just need to get the job done on the disease and be safe
to use. In fact, many drugs act on multiple targets as part of their
therapeutic mechanism.” I have frequently heard people make comments such as, “this
is just a probe, not a drug,” but a good probe should actually be more
selective than many drugs.
That said, you do want a drug to
actually hit the target of interest. The researchers highlight iniparib, a putative
PARP inhibitor that made it all the way to phase 3 clinical trials for breast
cancer and was tested in >2500 cancer patients. It failed. Moreover, that
failure cast a pall over the field which likely delayed the development of actual
PARP inhibitor drugs.
The researchers also discuss
aggregators, which are still being reported uncritically in the literature,
along with PAINS. “Such compounds should never be considered further or used as
chemical probes. They should be excluded from compound libraries. Yet many are
sold by commercial vendors as chemical probes and widely used.”
This statement raised the
hackles of Pete Kenny. In a recently published critique, he states: “it is asserted
in the conversation that commercial vendors are selling compounds as chemical probes
that are unfit for purpose and I strongly recommend that anybody making such
assertions should carefully examine the supporting evidence.”
Dear reader, please try the
following experiment. Enter “iniparib supplier” in your favorite search engine
and see what comes up. For me, the first 10 results include several that
describe it as a PARP inhibitor. I won’t link to them here because I don’t want
to encourage traffic to their sites. (This is also part of the reason Practical Fragments has discontinued
PAINS shaming, as it only increases the profile of sloppy or harmful
papers.)
Pete goes on to write: “I would
strongly advise against making statements that a compound is unfit for use as a
chemical probe unless the assertion is supported by measured data in the public
domain for the compound in question.”
Frankly, I don’t understand Pete’s
position, which I parodied here. Life is short and biology is complicated, so why
waste time with dirty or inadequately characterized reagents? For me, everything is an artifact until proven
otherwise. And the Chemical Probes Portal goes a long way towards demonstrating
whether a particular probe is fit for purpose.
Posts
9 comments:
Hi Dan, my position (“chemical probes will have to satisfy the same set of acceptability criteria whether or not they trigger structural alerts”) was articulated clearly at the start of the blog post and I’d argue that it’s actually very close to your position (“everything is an artifact until proven otherwise”). If people are flogging compounds as chemical probes then it’s perfectly acceptable to criticise them if they’ve not provided the evidence required to demonstrate that the compounds are indeed suitable for use as a chemical probes. What I don’t think is acceptable is assert that a compound is unsuitable for use as a chemical probe purely on the basis that it shares structural features with compounds (of undisclosed structure) that showed frequent-hitter behavior in a panel of 6 AlphaScreen assays. As a general point, structural alerts such as PAINS filters that are derived from screening assay results should only be used to assess screening results (and not much higher affinity ligands that have been subjected to careful evaluation).
I was just about to write that you two (DE & PK) are basically writing about the same elephant (in the room), just from different angles. To be clear, selectivity is a hard quality to prove, as it is costly and often assay dependent. One can only hope that true chemical probes emerge at the end of this confused tunnel. On a personal note, I really wish we could pool enough resources to settle this once and for all (CEREP Bioprint style, N compounds x M assays, full matrix, publicly available).
Hi Pete,
I don’t think Arrowsmith and Workman are suggesting that “a compound is unsuitable for use as a chemical probe purely on the basis that it shares structural features with compounds… that showed frequent-hitter behavior.” Indeed, in the passage you quoted they write that “simple lab assays should be run to check for the various problematic properties we mention.” At the same time, certain moieties do tend to cause problems in various assays. These are not always recognized, and as iniparib and many other examples demonstrate one cannot rely on commercial vendors to vet underlying mechanisms. Anyone using a compound should be aware of all the potential liabilities. That’s why the Chemical Probes Portal is so valuable.
Hi Tudor,
I agree that Pete and I do share some common ground. My perception is that Pete is more willing to entertain molecules as probes even if they contain dubious substructures, perhaps because of his discomfort with the original PAINS filters. I like your idea of pooling resources in a systematic way to investigate further. Sounds like a good grant proposal :)
My view, Tudor, is that the real elephant in the room is the uncertainly in intracellular free concentration (this will be the subject of a future blog post). My assessment, Dan, is that Cheryl and Paul do assert in the conversation that PAINS should never be considered further as chemical probes although I will concede that the criteria by which compounds can be validly labelled as PAINS are ambiguous. That said, in practical terms you’re never going to be in a situation where you make a decision as to whether a probe is acceptable on the basis of a substructural match. I’m assuming that the problem with iniparib is simply that there is insufficient evidence to demonstrate its acceptability as a chemical probe and that the perceived nastiness of structure is a red herring. My view is that structural alerts are more appropriately used to assess screening hits (e.g., > 50% inhibition at 10 μM) than optimized chemical probes.
Hello all
Let me try to be clear on my own position on the debate and also add a few other comments. Firstly I should say that I completely agree with Peter that otherwise high-quality chemical probes (exhibiting high potency and selectivity against the target, clear evidence of target engagement and modulation in cells, and also – for in vivo use eg in rodents – adequate PK/PD and tolerability) should not be deemed unsuitable purely on the basis that they share structural features with compounds that exhibit frequent hitter behaviour. As Cheryl and I said in the Nature Communications conversation piece, and Dan quotes us on, we think that if there is a concern, based on a structural alert, then the ‘simple lab assays’ we refer to should be used to check whether any suspected behaviour is actually exhibited. Assays are more important than predictions. A general comment I would make though is that the Chemical Probes Portal was created and developed mainly to provide advice to the majority of biologists, most of whom are not experts in chemical probes and the issues that are being debated in this exchange of views, who want to use chemical tools in their research. Although experts in chemical biology and drug discovery also find the Portal useful, it is this majority of biologists, who might otherwise inadvertently choose and use very poor quality compounds – such as the aforementioned iniparib, gossypol, curcumin and many many others – as chemical probes, who can most benefit from the advice provided. In view of the continuing use of poor-quality compounds as chemical probes we do feel that is far better in our advice to err on the side of caution. I think that the maxim that the effects of all compounds in biological systems should always be considered to be artefacts until proven otherwise is a good one to follow. To Peter: we will be revising the information we provide on Portal about PAINS alerts and toxicophores to make our position clearer and we are also looking at how we present and update the list of the ‘Historical Compounds’ that Dan refers to and which include both seriously flawed nuisance compounds and also a number of compounds that may have been useful in the past as pathfinder compounds, but which have been superseded by higher-quality chemical probes. Users of the Portal consistently tell us that the Historical Compound list is very helpful. We appreciate comments from the user community that we serve. I think we would all want to see the elimination of the most egregious compounds from use as chemical probes in biology, while also striving to develop increasing numbers of high-quality chemical probes across the proteome and promoting their best-practice use in biological research. With the recommendations being based on measured data on chemical and biological properties.
Hi Paul,
Thanks for your excellent comments. I like the idea of separating the Historical Compounds into two lists, one of which includes older or flawed probes while the other includes true nuisance compounds. Perhaps you could call these "False Probes" or, channeling the Baell & Walters Nature Comment, "Con Artist Probes."
Hi Peter,
Contrary to your assumption that "the problem with iniparib is simply that there is insufficient evidence to demonstrate its acceptability as a chemical probe," iniparib has actually been invalidated as a chemical probe; see for example reference 10 in the Arrowsmith & Workman Conversation. Specifically, it does not inhibit PARP enzymatic activity. Despite the fact that reference 10 was published in 2013, suppliers continue to sell iniparib as a PARP inhibitor.
Thanks to you Dan for highlighting iniparib as good example of a compound that is still sold by vendors as a chemical probe years after it was invalidated both by showing as you say that it does not inhibit to the claimed target (PARP) and also by demonstrating that it exhibits the promiscuous chemical reactivity with proteins that might be predicted from its structure. In addition by the way, it failed in the Phase III clinical trial in breast cancer patients. Thanks as well for your positive comments on the developments of the Chemical Probes Portal, as highlighted in our Nucleic Acids Research data base issue paper.
Thanks for your comments and clarification, Paul, and I should stress that my criticisms of the way that structural alerts are used to assess chemical probes are all to do with the associated cheminformatics. My view is that cheminformatic models for prediction of frequent-hitter behavior, even when based on larger (many assays) and more diverse (many assay read-outs and targets) data sets than that used in the original PAINS study, provide minimal guidance as to which confirmatory checks should be carried out. I would argue that labels such as “nuisance compound” should only be used of supported by relevant measured data that is available in the public domain. Promiscuity and read-out interference both increase with the concentration of the compound and one needs to be careful using the term “promiscuous” in the context of chemical probe quality (I prefer the term “frequent-hitter” when compounds have only been assayed at a single concentration).
Hi Dan, I wasn’t aware of the extent of the dirt on iniparib so thanks for flagging this up. My view is that if there are insufficient measured data in the public domain to support the use of a compound as a chemical probe then this should be flagged up and I see no need for predictions from questionable cheminformatic models. Although many PAINS substructure look thoroughly nasty (I’d seen many of them while I was still at AstraZeneca) it’s worth remembering that the assay panel used in the original PAINS article consisted of 6 assays for protein-protein interactions (no catalytic cysteines) that all used the AlphaScreen read-out and I consider quenching of (or reaction with) singlet oxygen as the most likely explanation for frequent-hitter behavior observed in this proprietary data set.
Post a Comment