Skip to Content

Those Compounds Aren’t What You Think They Are

There’s a lot of work in the literature on the TrkB receptor, which responds to brain-derived neutrotrophic factor (BDNF). That name certainly makes the ligand protein sound like a pretty big deal, and so it is: BDNF is involved in a lot of neural development pathways, injuries to nerve tissue, and the like, and given the black-box nature of so many of those processes, it and the other neurotrophin hormones have been the subject of a great deal of research. There’s an oncology angle as well, since fusion proteins involving BDNF and others have been found apparently driving some types of cancer.

That means that there’s potentially room for both agonists and antagonists of the receptor as therapeutic options. Antagonists are being looked at in oncology, and agonists could be of use in Alzheimer’s or other neurodegenerative conditions. But coming up with an agonist of a receptor like this isn’t so easy – the whole class of G-protein-coupled receptors that respond to big protein ligands is notoriously hard to deal with in that fashion, replacing their natural partner with a small molecule, because the binding surfaces are rather large and complex. You’d think, well, fine, I’ll just infuse BDNF right into the brain, because these are desperate conditions that require desperate measures. But every single attempt to use that as a therapy has (to my knowledge, failed) so there’s a lot going on here.

Nonetheless, there are several TrkB agonist compounds reported in the literature, and they’re rather small molecules (mostly flavone derivatives). 7,8-dihydroxyflavone in particular has had a lot of work done on it, in numerous models of disease. It does seem odd that a molecule this size can activate the TrkB receptor, and flavone derivatives in general are known as a class of compounds whose activities are hard to pin down. There are other structural classes reported in this field as well, though, such as LM22A-4 and the classic CNS molecules amitryptaline and deprenyl.

Doubts have been raised about these compounds, although publications on them seem to be continuing at a steady pace. But a new paper throws the gauntlet down. The authors, from Columbia and the Broad Institute, have taken a more detailed look at receptor phosphorylation (which is the big event on activation of this class) and downstream readouts (AKT, ERK1, ERK2, etc.), and have come to the conclusion that none of the reported TrkB agonists are, in fact, TrkB agonists at all. BDNF and neurotrophin-4 light up these assays in the just ways that you’d expect, but not the small molecules. The paper also describes a screen through 40,000 compounds in the Broad’s diversity-oriented synthesis collection, which in the end led to no confirmed hits, either. Finding a small-molecule ligand for this receptor is indeed not an easy task.

So why have these compounds persisted in the literature as long as they have? The authors suggest that the methods involved have obscured things:

We propose that one plausible explanation for the observed discrepancies is methodological in nature. Although Western blotting is a standard core method in life sciences research, it requires many procedural steps, making it impractical to perform multiple repeats for each experimental condition plus controls (4749). Moreover, densitometry is dependent on the analysis procedure, which may lead to biased results, especially with a low amount of phosphorylation of the protein of interest (50). Specificity of the antibody plays an important role as well (5051). Although antibodies recognizing phosphorylated ERK and phosphorylated AKT yielded highly specific staining on blots (fig. S5), we did not find an antibody that was specific for phosphorylated TrkB; the commercially available antibodies tended to recognize additional targets. Complementary methods should be applied to confirm Western blot results. In our view, the described quantitative ELISA provides sufficient throughput for controls and repeats to be used along with the blot to more accurately characterize drug activity. We recommend that Western blot be used as a qualitative or semiquantitative complementary method to the quantitative ELISA assays, as confirmation that ELISA indeed detects the desired phosphorylation reaction. Alternative methods that are independent of antibodies should also be considered, such as methods based on enzymatic activity of reporter proteins as the readout.

When you consider the number of reported results (not just in this field!) that rely on Western blots, this is food for thought. (And once again, antibodies that aren’t as good as they’re supposed to be cause trouble). It’s not a new worry, by any means, but you really do want to keep this in mind when you’re evaluating published work (and keep in mind that these worries are completely separate from ones about outright Western <s>blog</s> blot fakery, which has shown up many times in the last few years). A lot of time and money has been spent on these compounds, and a lot of hypotheses generated which are now going to have to be re-evaluated, to say the least.

7,8-dihydoxyflavone may well have effects in animal models, but they’re apparently independent of anything involving direct activation of TrkB. There could well be indirect TrkB mechanisms (as the authors of this paper take time to point out), involving things that aren’t present in many of the in vitro assays. But that’s not what people have been working on the basis of, and if such mechanisms are indeed in effect, time would have been better spent tracking them down rather that working on a false basis. Let’s see how this paper goes over in the field. . .

30 comments on “Those Compounds Aren’t What You Think They Are”

  1. Mftkoehler says:

    “… Western blog fakery…” is a priceless Freudian slip!

    1. Derek Lowe says:

      I have to agree – I’m not completely deleting it!

  2. Lars says:

    N-acetylserotonin (an intermediate in the melatonin biosynthesis) seems to be a TrkB agonist. I’m surprised you don’t mention this class of compounds? https://www.ncbi.nlm.nih.gov/pubmed/24664769

  3. Lane Simonian says:

    Even if you could find an agonist for TrkB receptors it would not likely help in the treatment of Alzheimer’s disease. Nearly every g protein-coupled receptor is damaged in Alzheimer’s disease (likely due to cysteine oxidation). This includes receptors for smell and receptors for neurotransmitters involved in the retrieval of short-term memory (acetylcholine), sleep (melatonin), alertness (dopamine), social recognition (oxytocin), and mood (serotonin). The one exception appears to be receptors for noradrenaline; and high levels of noradrenaline are suspected to be the cause of neuropsychiatric problems in Alzheimer’s disease.

    It is unlikely that any agonists for any g protein-coupled receptor will help in the treatment of Alzheimer’s disease unless cysteine oxidation is reversed first.

  4. Isidore says:

    “The authors, from Colombia and the Broad Institute […]”
    It should be Columbia, the university, and not Colombia, the country.

    1. Derek Lowe says:

      Dang it. I do that more than random chance can account for.

  5. roger says:

    “densitometry is dependent on the analysis procedure, which may lead to biased results…”
    The shade of A. van Maanen smiles!

  6. Umed Boltaev says:

    I would like to point out that TrkB is part of receptor tyrosine kinase (RTK) family, not G-protein coupled receptors (GPCR). These are 2 very different receptor family, RTK have single transmembrane helix and are activated through dimerisation (oligomerization), while GPCR are mostly membrane proteins and have multiple transmembrane helixes and are activated mostly through confirmational changes upon ligand binding. There are no good example of RTK agonist, while GPCR had many successes. The main reason of this difference between 2 family in drug development could be attributed to their structures: GPCR are usually activated by small molecules and have hydrophobic ligand binding domain, RTKs require protein as a ligand and doesn’t contain any domain that could bind small molecules (except for kinase domain). RTKs are referred as undruggable in the scientific community.

    1. ScientistSailor says:

      That’s quite a statement. I’m in the scientific community, and have never heard of RTKs referred to as ‘undruggable.’ There are antibodies that activate MET receptor through dimerization. They can also be drugged through inhibition of the kinase activity, and there are several drugs on the market that do this, not what I would call ‘undruggable’.

    2. Derek Lowe says:

      That’s all true – I didn’t go into in it the post, but the RTKs are indeed different beasts, and very hard to target. Which makes the reports of small-molecule agonists all the stranger. . .

    3. Lane Simonian says:

      Thank you, unmed Boltaev, for pointing out that the TrkB receptor is a receptor tyrosine kinase and not a g protein-coupled receptor. The curious part of this is that overactivation of receptor tyrosine kinases and/or g protein-coupled receptors is an early event which can lead to Alzheimer’s disease.

      Overactivation of these class of receptors leads to cysteine oxidation and tyrosine nitration. Tyrosine nitration can inhibit the neuroprotective phosphatidylinositol 3-kinase/Akt pathway. Along with the cysteine oxidation of g protein-coupled pathway, tyrosine nitration plays a critical role in the onset and progression of Alzheimer’s disease.

      https://www.ncbi.nlm.nih.gov/pubmed/9826526

      https://www.ncbi.nlm.nih.gov/pubmed/16410804

      https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3645702/

  7. peptoidchemist says:

    oh hello Dalibor Sames

  8. AC says:

    Anyone want to place bets on how many papers will continue to use and cite these molecules as agonists?

    This brings to mind how rottlerin is still being used as a PKC inhibitor despite it being shown that it is not. Pubmed says that there have been about 300 papers published still using it as such, despite a 2007 paper called “Rottlerin: an inappropriate and ineffective inhibitor of PKCdelta” .

    http://blogs.sciencemag.org/pipeline/archives/2015/04/01/rotten_rottlerin

    1. Derek Lowe says:

      Unfortunately, you’re probably right. Rotterlin’s a good (bad) example of how these things take on a life of their own. . .

  9. milkshaken says:

    I worked on a medchem project trying to identify small molecules that have BNDF-like effect on retinal ganglion cells (these are specialized neurons that die out in glaucoma). We found several old kinase inhibitors (Sutent, SU-6656, VX-680) that had a BNDF-like neuroprotective effect. These were poorly selective kinase inhibitors, and were hitting TrkB-related kinase targets somewhere along the signaling path; they were not receptor ligands. Also the original targets of there inhibitors had absolutely to do with BNDF-like activity. When you have Sutent-like ligand that inhibits one third of human kinome below 100nM, puzzling results are to be expected

    1. tangent says:

      Sunitinib, man. If I understood why it is not highly toxic, maybe I would know the first thing about kinases.

      1. Belgian PhD student says:

        Biodistribution perhaps?

        1. milkshaken says:

          Sutent is actually pretty toxic in animals, the tox in dogs was quite horrendous (adrenal necrosis, organ accumulation) and it has long lived cardiotoxic desethyl metabolite that lingers and causes QTc prolongation. The therapy includes a drug free “vacation” for the patients, to stop the buildup.

  10. Colombia University, Bogota says:

    You have to hand it to Dalibor Sames — his reputation was destroyed in the organometallic community for the 5+ fake papers by his student Bengu Sezen (that he defended) so he reinvented himself and did neuro drug discovery.

  11. hn says:

    Don’t trust any biological results that are based only on Western blots.

  12. Signaling guy says:

    If you want two _real_ examples of small molecules that can activate cytokine/growth factor receptors please see below (granted, these are not RTK’s, but these cytokine/growth factor are equally challenging to target based on first principles). Both were identified by scientists at Ligand Pharmaceuticals in the late 1990’s. Note that the second example led to an approved GSK drug, now owned by Novartis (eltrombopag or Promacta):
    First example:
    A small, nonpeptidyl mimic of granulocyte-colony-stimulating factor
    Tian SS et al. Science. 1998 Jul 10;281(5374):257-9.
    Selective binding and oligomerization of the murine granulocyte colony-stimulating factor receptor by a low molecular weight, nonpeptidyl ligand.
    Doyle ML et al. J Biol Chem. 2003 Mar 14;278(11):9426-34.
    Second example:
    Discovery and characterization of a selective, nonpeptidyl thrombopoietin receptor agonist.
    Erickson-Miller CL et al. Exp Hematol. 2005 Jan;33(1):85-93.
    NMR structural studies of interactions of a small, nonpeptidyl Tpo mimic with the thrombopoietin receptor extracellular juxtamembrane and transmembrane domains.
    Kim MJ et al. J Biol Chem. 2007 May 11;282(19):14253-61

  13. Signaling guy says:

    It’s not impossible. If you want two _real_ examples of small molecules that can activate cytokine/growth factor receptors please see below (granted, these are not RTK’s, but the cytokine/growth factor receptors involved are equally challenging to target based on first principles). Both were identified by scientists at Ligand Pharmaceuticals in the late 1990’s. Note that the second example led to an approved GSK drug, now owned by Novartis (eltrombopag or Promacta):
    First example:
    A small, nonpeptidyl mimic of granulocyte-colony-stimulating factor
    Tian SS et al. Science. 1998 Jul 10;281(5374):257-9.
    Selective binding and oligomerization of the murine granulocyte colony-stimulating factor receptor by a low molecular weight, nonpeptidyl ligand.
    Doyle ML et al. J Biol Chem. 2003 Mar 14;278(11):9426-34.
    Second example:
    Discovery and characterization of a selective, nonpeptidyl thrombopoietin receptor agonist.
    Erickson-Miller CL et al. Exp Hematol. 2005 Jan;33(1):85-93.
    NMR structural studies of interactions of a small, nonpeptidyl Tpo mimic with the thrombopoietin receptor extracellular juxtamembrane and transmembrane domains.
    Kim MJ et al. J Biol Chem. 2007 May 11;282(19):14253-61

    1. sgcox says:

      Eltrombopag is a one hell of the molecule.
      Would be interesting to hear how Ligand Pharmaceuticals made GSK and Novartis to take on it.
      If it comes from an internal screen (target of phenotipic – no matter) it would be dismissed as non-specific metal chelator, reactive, unstable, etc. All is acually correct but it is a marketed drug !

  14. Not Boston says:

    I’ve worked on this – the PLOS paper pretty much showed this – no idea how this is news. There’s nothing here – Derek I hope you aren’t featuring this paper because it’s in the Science family?

  15. DrOcto says:

    Thinking out loud here.

    But if RTKs are activated by dimerisation/oligerisation, presumably due to a large signalling protien that brings them together, then a good small molecule ligand could never act as an agonist. However if you did have a strong binding ligand, you could attach several of those ligands to an inoccuous scaffold/dendrimer. Might have the same effect.

    What is known about the binding domain(s) of BDNF? Is there any sort of repeated structural motif that might be indicitive the above?

  16. Jim says:

    Greatest scientific insight of the 21st century: unconscious manipulation. 99.9% of everything you say is hidden attempts at manipulation of others toward your selfish goals ( usually sexual or power related). This includes all of science, art, politics, music, etc. All of this is predicted by evolutionary theory.

    1. drsnowboard says:

      It’s life, Jim, but not as we know it… or rather, no life as we know it?

      1. Jimmy John freak show says:

        Not sure what that means but its manipulative for sure…and evil.

    2. DrOcto says:

      I hear 86.3% of all statistics are made up on the spot.

      1. Jimv2.0 says:

        If i were to say 100%, it would be a theory, and no longer a stat……OK, 100% then.

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.