Submission MAT-POS-090737b9

Topic automatically created for discussing the designs at:
https://covid.postera.ai/covid/submissions/MAT-POS-090737b9

Hi @mc-robinson @londonir @alphalee @JohnChodera @frankvondelft

In order to give yourselves the best chance of making this approach work, I’d recommend that you also look at the 3-chlorobenzyl analogs such as VLA-UCB-05e51b3f-16 and DAR-DIA-56cf811e-1 since the additional rigidity of the chromane may well destabilize the transition state. The 4-chlorobenzyl analog PET-UNK-e8c7a26f-1 may also be worth looking at since SAR does not always track between series of reversible and irreversible inhibitors. This commentary on the PET-UNK-e8c7a26f submission may be helpful. I’ll also mention @JSPEN @AnthonyA @Daren_Fearon with whom I’ve had discussions on this topic and @edgriffen @RGlen from the design team.

1 Like

We’ve had real formation. isolation/purification problems, with similar analogues I’m afraid though Enamine managed to make one or 2, but I think in fairly low yields and I’m not sure that they’re that stable… if anyone had a nice recipe to follow, please share as we submitted warheads in easier to make scaffolds I’m afraid.

Hi John, does this mean that the synthetic chemistry works with the chromane but not the 3-chlorobenzyl?

Hi @JSPEN, indeed Enamine has made this with the 6 membered ring pictured and the 6,5 bicyclic system (with the benzofuran like scaffold). Unfortunately, stability was a bit issue there and we barely got the few mg needed for testing. It’s actually shown good MPro activity, but the antiviral activity is of question and we saw a bit of cytotoxic. I think the stability of that imide like warhead scaffold is of concern. I wonder if we can find something other than adding acrylamide at the last step that replicates the warhead geometry with better stability.

@mc-robinson @londonir @pwkenny hello! I think perhaps an e.g. RCO(NCOCH=CH2)R1 type structure, whatever the R groups, is unstable. Maybe, someone could calculate these? Are the CO’s destabilising one another? I saw a few recent Nir London submissions with similar motifs e.g. NIR-THE-af15c15d ? Have they been made? I made a few suggestions e.g. placing the warhead away from the amide N, possibly on an aromatic scaffold e.g. JOH-UNI-3fc3434e? This would need to be modelled. Using a hydrazine to have RN(NHCO-CH=CH2)COR’ would space it a further heteroatom off the amide but we struggled to selectively amide couple hydrazines too, e.g. JOH-UNI-ea72002d. Perhaps diazotizing an amide and reducing is a better route that we should try.
It’s one of those dilemmas where you struggle to make something that’s also quite unstable, so you give up as it’s difficult to envisage it as a potential drug or tool.

Hi John,

I recall your earlier concerns about stability and apologies for not having responded properly at the time. I agree that we do need to get a handle on this issue since it’s likely to be very difficult to make compounds more stable without compromising activity. I’m not sure how reliable calculated stabilities would be and I’m not in a position to do QM calculations. That said, if someone has access to QM software then I’m happy to discuss options. The amidic nitrogen in these structures is attached to two carbonyl groups and a moderately electron-withdrawing pyridine/isoquinoline which I’d expect to make them better Michael acceptors and more vulnerable to hydrolysis (more ester-like than amide-like) than a typical acrylamide.

If using a hydrazine linker, I’d recommend attaching the electrophilic center directly to nitrogen and I previously submitted the hydrazine-linked nitrile PET-UNK-a692de38-1. I’d expect a hydrazine linker to not invert the cis/trans preference of the amide as would an sp3 carbon linker (see notes that were added as a comment on the original submission). I’d also be thinking about the N-methyl analog since a fully substituted hydrazine with an electron-withdrawing group on each nitrogen doesn’t look too much like a hydrazine. Azapeptide nitriles can actually be more potent cysteine protease inhibitors than the corresponding peptide nitriles. This comparative study of warheads and the PDB:6P4E X-ray crystal structure (here’s link to article since that’s not included in PDB entry) may be of interest.