Kinase Chemistry – Just a year and a half behind the times.


Posted by kinasepro on May 17, 2007


Supergen just gave an update in Italy yesterday: webcast here. Skip to ~7 minutes for the kinase portion. The highlight for me is on MP-470 where he says that its uM on met (and ret).

but at AACR last year they suggested that:

In their defence I’ll postulate that they meant 4,000 nM concentrations last year. Ultimately does it matter? Only in so far as how much you can trust what they say… heh


16 Responses to “MP-470”

  1. milkshake said

    I can’t believe somebody would even present a micromolar compound, repeatedly, and at a high-profile meeting.

    We get decent 1 uM compounds all the time from our high-throughput screening, 3-5 reasonable dissimilar structures per target screened. But the amount of work that one needs to do, to turn it into anything close to a preclinical candidate is usually something over 10 man-years.

    Perhaps they presented nanomolar biochemical potency and later talked about data from cell-based assay in their business presentation.

    Also biochemical assay can be tweaked, to produce low numbers for presentation purposes.

  2. Yeah milkshake, I was just thinking that perhaps they increased the concentration of ATP in one of the assays to increase the IC50, or that one value refers to a cell-based IC50.

  3. kinasepro said

    It’s still a mutant c-kit inhbitor with pdgfr and flt3 activity, which make it an interesting compound, its just not what you and I would consider a met inhibitor.

  4. weirdo said

    I agree that presenting a micromolar compound is ridiculous. It’s academia-level medchem.

    But I’d rather start with a micromolar lead with good PK properties than a nM lead with crap bioavailability any day of the week and twice on Sunday. Taking a micromolar hit to a nM lead is ridiculously easy — by that I mean you can either do it in a few months, or know enough to quit. But building in oral bioavailability to a lead that isn’t? No, thank you.

  5. Hari said

    Med.Chem guys are overconfident with nM all the time and end-up in wasting time in making 100’s of compounds. STI-571 is a mM inhibitor for Bcr-Abl.
    They should think about not just make pico or nM inhibitors ,also very important at early stage to confirm its T 1/2 and % bioavailability. Correct?.
    MP-470 is in Phase I now.

  6. While high potency in terms of nM IC50 is definitely important, here’s something that Ronald Breslow said in an paper about the development of an inhibitor derived ultimately from DMSO:

    “It is probably a mistake for medicinal chemists to set out first to find the most potent compound they can achieve in a series and then to look at any question of toxicity, as is often done. As with SAHA (the compound that Breslow and his colleagues discovered), in other series there will be examples in which the medium potency compound has enough efficacy to be a useful drug but not so much as to cause unacceptable side effects.”

    More simply, Breslow is saying that high potency compounds may also be high toxicity compounds because they are tight binding and keep the target modulated for longer than what’s desired. So it may make sense to focus on intermediate range potency and pursue other properties such as solubulity, bioavailability etc. for these compounds, as Hari mentioned above.

  7. weirdo said

    Yeah, I’m going to take drug discovery advice from that great medicinal chemist, Ron Breslow.


    Potency matters. It doesn’t matter MORE than any other parameters, but it matters. More so for some targets than others.

    Also, if you are implying that high potency is somehow related to the kinetics of the binding event, that is clearly not so. Ki need not be correlated to the off-rate. I’m sure Breslow knows this, so I think you are mis-interpreting him.

    If you need a 1 gram pill and micromolar blood levels to achieve efficacy, you’re going to have a safe drug? Not bloody likely.

  8. kinasepro said

    The IND was accepted by the FDA triggering a $10 million milestone to Montigen’s investors. They haven’t announced a ph1 yet though.

    Ultimately blood levels, and protien binding, and herg, and target profile and chemotype and and and and… You know, it really doesn’t matter. Nope not a bit, none of it. All that matters is safety and efficacy – period.

    To me what does matter though, is when a company jumps up and down about something (like – c-Met activity) and then it turns out they were misinformed. Can you say – amateur hour?

  9. Weirdo, I never said that potency does not matter. But what Breslow is trying to say is that always shooting for the most potent compound may have problems of its own, and sometimes going for an intermediate compound may turn out to be a more effective strategy. You are exaggerating with the 1 g example!

    Also, there are surely cases where potency is related to the K (off). Here’s what Breslow says about it:

    “With ligands to biological receptors, there is seldom, if at all, any good way to increase the on rate, so high potency— strong binding—is always reflected in the off rate. That is, as compounds are more strongly bound they are released from the binding pocket more slowly in a first-order processunaffected by the concentration of the free ligand.”

    Again, this cannot be completely general but I don’t see any reason why it should not be regarded as a possible reason for problems with high-potency compounds.

  10. h2l said

    KP. We happened to meet with these guys at AACR and I think this compound is interesting but I would not call it a specific c-Met inhibitor. From what they showed us, It has decent biochemcal activivty (30-50nM) against c-Kit, PDGFR and Flt-3 and although it was unclear to me what the activity was for c-Met they did show cell and xenograft data showing modulation of c-Met activity at high nM to low uM range. This compound has a very interesting tox profile and they showed sustainable low uM plasma levels with modulation of PD markers and efficacy with little tox. Just like every other compound at this stage, we will see what happens in the clinic. I don’t think the compound is bad the communication seems a little suspect.

  11. David said

    Just look at the structure, there are two potential issues associated with the molecule. One is the thio urea, the other is the dioxy moiety.

    In general, if the toxicity is mechanism based, more potent compound don’t help, but if that’s off target, I think more potent compound with similar other properties such as protein binding, herg and solubility will translate lower dose, which will have less tox issues.

  12. milkshake said

    I can tell you how this molecule most likely binds: The methylene dioxo is a hydrogen bond acceptor that goes deep into a narrow binding pocket next to the ATP binding site. The tricyclic goodness with the thiourea piperazine is pi-stacking (mostly by its electron-deficient pyrimidine ring) with the aryl on tyrosine side-chain in the activation loop (that is normally invisible in X-ray because it is disorganized) and by doing so it effectively puts a lid over the active site, binding perpendicularly with respect to a plane of the classical ATP-site binders go.

  13. kinasepro said

    Hmm… I’m not sure I get your drift. Words are worse then pictures, but sadly the picture they show in the poster is bad. Really bad.

    It should be no secret by now that I think the best way to figure out how people think about their own inhibitors is to look at the other compounds they claim, and sure enough one finds in WO/2005/037825 that the tricyclic moeity appears to be a quinazoline isostere, with the thiourea heading to H1.

  14. milkshake said

    If the terrible tricycle is quinazoline mimic then I am wrong and it is a classical binder. I got led to these speculations because of some striking similarities with some other c-Met compounds whose cocrystalls I saw

  15. OrgMed said

    Hello Kinase Pro: You are right I have no clue what milkshake trying to say on the binding mode. The Benzofuranopyrimidine moiety is binding at ATP adenine moiety, the piperazine involved in interactions with sugar binding residues and the methylenedioxy moiety extended into alpha phosphate binding region.

  16. kinasepro said

    Thanks OrgMed. I’d love to see it in 3D. Tell Supg to submit it to the PDB plz! The phosphate binding region isn’t H1 though, and is more I think what Milks is alluding too. As for c-met… Well, with it being micromolar, I’d say nothing in the rulebook precludes an alt binding mode.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: