Suggestion for N-substitution on the piperazine ring

[mattodowd25]

Hi MyOS community - my name is Matthew and I am a current honours student in Mat’s group. Over the next few months I will be working alongside the community to contribute compounds to the library. (edit by @mattodd: Matt's ELN is now online)

The current issued faced by the MyOS project is which compound to synthesise next? Several sources of molecules already exist. These include a library of drugs screened from a Perth based medicinal chemistry company, Epichem, and from a previous honours student, @fantasy121 . All data about their efficacy and pharmacokinetics can be accessed through the master list, as a separate sheet. With your help, it may be possible to screen the data to determine an active scaffold that can be modified into a lead candidate.

One structure I am particularly interested in exploring is pictured below. It has been noted in the preprint as an active structure. My major aim over the coming months is to vary the R group, determining the extent of its activity.

Therefore, it is essential that the data in the master list be screened, to see which compounds I might make are already in the Epichem library and therefore don’t need to be synthesised. The data on all the Epichem compounds (actives, inactives and the list of all the compounds not yet screened) can be found in https://github.com/OpenSourceMycetoma/Series-1-Fenarimols/issues/1. Analysing the data may provide ideas for new R groups. I am currently in the final steps of the synthesis towards the Boc protected intermediate (see below). The question is which R group does the community believe will be of greatest benefit to synthesise based on the available data? (https://github.com/OpenSourceMycetoma/Series-1-Fenarimols/issues/2)

[MFernflower]

Might I suggest reacting the piperazinyl core with methyl chlorothiolformate and ethyl chlorothiolformate ? I know these are very ferocious chemicals but the thiocarbamates made might be of use (Being heavier Atom analogs of the good carbamate compounds) @mattodowd25

[mattodowd25]

@MFernflower good suggestions. I would recommend checking out these two compounds in the EpiChem library, EPL-BS0592 & EPL-BS0595. There is also EPL-BS0690 which has been screened against mycetoma. All have a modified thiocarbamate. I believe all are incorporated into a heterocyclic ring.

[MFernflower]

Those are not really thiocarbamates per se - Here are some possible ideas I came up with

@mattodowd25

@mattodd All compounds in this post are not in the epichem library

[mattodd]

I think what we need here is a scheme of all compounds based on this structure with N-R (i.e. where N is currently (in the lab) Boc, precursor to EPL-BS0495 in #2, HPD12_1 in Master List) that are in the Epichem library. We need to know which have been evaluated (and whether active (green) or not (red) - structures are in #1) or whether they have not yet been evaluated (black). @mattodowd25 can you please generate this scheme and paste below as a PNG file (and as a .cdx file, though you'll need to zip that for Github to accept it)?

This will allow us to be sure that any proposed compounds (or similar structures) are not already in the Epichem library. For any proposed instances of R, can people please be sure to check SciFinder or Pubchem for anything similar (90% Tanimoto for example), again so we can be sure that the compounds actually require synthesis. Boc Variants.zip

[mattodowd25]

Attached is a scheme outlining the compounds that are known with the scaffold outlined above. In the scheme below green is active, red is inactive (edit by @mattodd - no, red is "currently under evaluation", as per the legend), and black not yet tested. Before suggesting new compounds make sure that they are not currently on SciFinder, or PubChem at 90% Tanimoto. Compounds

[MFernflower]

@mattodowd25 @mattodd @drc007 Compounds similar to a few of the untested ones in that PDF have been screened but there is a problem - we do not know if the N-substitution killed their potency as the F and Cl atoms were swapped for other things per each molecule (please excuse the mspaint graphics)

I can infer that the methyl 1,2,4-oxadiazole motif is useful in these drugs and that piperazinylmethylsulfone is not (see first two compounds)

As for a priority compound we should try bolting that methyl 1,2,4-oxadiazole motif onto our newly made core! note: the precursor is quite costly - http://www.matrixscientific.com/065590.html A related compound could however be made quite cheaply - https://www.sigmaaldrich.com/catalog/product/aldrich/748250?lang=en&region=US (I have a hunch this should work fine as I bet that smaller ring is just binding the iron center of cyp51 - the unpaired electrons of the aromatic N atom interact with the iron)

PDB: 3JUS

[mattodd]

Hi @mattodowd25 - so that's four relevant compounds already in the library, great. Could you re-draw the scheme to remove the duplication - i.e. have the consistent fragment up top, and then just the variable parts sketched out? Like this. That way it should fit on one page. Convert the new chemdraw to a png and drag it onto your post above (so we can see the pic directly), then zip the .cdx file and drag that too, so that we can download and play with it.

But the upshot is: some available compounds. @bendndi - would you be amenable to a screen of these, given the promise of the two analogous compounds evaluated thus far?

@mattodowd25 for the ones in Pubchem, what are the sources? Single reports, or multiple?

@MFernflower - agreed about the differences for those other compounds, though thanks for pointing them out. The as-yet-untested compounds would give us new data for the SAR, and there's already some compounds with 5-membered heterocycles in there. What's missing and suitable for synthesis? Sulfonamide? Your cyclobutane suggestion looks a bit greasy to me. Any reason for the thiocarbamates?

[MFernflower]

The idea behind the thiocarbamates was that they were simply heaver analogues of the known active (ethyl) and inactive (methyl) - as for the pyridine and cyclobutane compounds - to be honest now that I looked at the binding pocket (see image above) I doubt they would work whatsoever and feel they should be scrapped!!!!

@mattodd Your concerns about greasiness - In general most all of the cyp51 inhibitor class is rather oily - this might even prove beneficial as @wwjvdsande reported the grains of the fungus being very lipid and protein rich - hence why I wanted to screen paclobutrazol against the fungus!

My vote is now to make a compound using this precursor: https://www.sigmaaldrich.com/catalog/product/aldrich/748250?lang=en&region=US - also the compounds in that list have a carbonyl linking the piperdine ring and the 5 atom cycle - these compounds are probably known inactives (the inactive carbonyl pentacycle compound did however have a p-cyano on the benzene ring so no clue if that could have killed potency rather than the carbonyl spacer in that same compound)

Compound to test if P-cyano kills potency: @mattodowd25

[MFernflower]

A urea might be fun to try:

[bendndi]

Urea is a good suggestion; there are none in the 800-strong Epichem library so it would be advantageous to make one and see it's impact. In same vein I would suggest making compounds of the following type via Reductive amination /alkylation / Buchwald; There are no examples of these types of modifications in the Epichem library so would be interesting to see what happens. also significantly change the physchem properties of the N of the piperazine, and in turn could have interetsting effect on in vivo deffect if the compounds are indeed active (see postulated correlation between logD and in vivo effeicacy in the biorxiv paper)

[MFernflower]

@mattodowd25 @bendndi Possible other idea:

Precursor chemical: https://www.sigmaaldrich.com/catalog/product/aldrich/d141208?lang=en&region=AU (32 AUD for 100 grams)

DPH is known to kill bacteria but antifungal effects of dph are still under research AFIK: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3768775/ @wwjvdsande do you know if any screens have been done where diphenhydramine is given to various fungi - My literature searching is pretty limited as I do not have access to paywalled papers

I think it's really important we try to cover large swatches of the chemical space around this compound

[mattodowd25]

I have attached a list of compounds that I would like to make of the coming weeks, or that I am currently synthesising. The morpholine sample should hopefully be purified by the end of the week. Also, the post made by me above containing compounds with this scaffold in the epichem library has been updated.

[MFernflower]

A close relative of the n-ethyl compound exists in the library as EPL-BS0167

@mattodowd25 @mattodd @bendndi

Unrelated - but I would be interested in seeing a compound like the morpholine one with the morpholine replaced with homopiperidine

[mattodd]

Wow, there are certainly plenty of structures that are available for further evaluation! @bendndi I guess we need to consider which of the compounds @mattodowd25 identified as existing in the Epichem library (above) we should screen. Is there any reason not to try them all, to fill out the SAR around the potent hits?

One compound, the one without the EPL code, appears not to be in the epichem master list (Matt searched for everything via SciFinder/Pubhchem) but is available from a DNDi/Epichem paper. I'll ask Epichem about this offline (they're not yet on this site).

People like the idea of a urea. There is one in the Epichem set (EPL-BS0447), so let's look at it. It's N-phenyl, so there could be a reason to make an N-ethyl or N-dimethyl as @MFernflower suggested?

I like all four suggestions from @bendndi involving benzyl, methyl acetate, methyl amide and pyridyl substituents. Pendant amide, ether and alcohol chains from @mattodowd25 also good. @mattodowd25 notice I've lengthened the alcohol and ether chains a little - your ether contained a nasty reactive aminal.

@MFernflower I agree about the close relative EPL-BS0167, but I'd still be interested in whether the F is important for activity, and given the ease with which this can be made, I think we could easily justify this synthesis. Your homopiperidine suggestion is interesting - what about simply piperidine as a presumed-negative control? Is this known, @mattodowd25 ?

So @mattodowd25 's list above is a great start, and we could then consider, in a second wave, the following. I've added ortho- and para-pyridines, but we should probably access meta- while we're at it.

@mattodowd25 can you briefly comment on how these proposed compounds compare with the Epichem compounds (the ones in black above that we're considering screening) in terms of calculated logP, and/or different physicochemical properties like molecular weight? Datawarrior might allow you to plot the two sets of compounds (i.e. existing Epichem compounds vs. our proposed molecules) in different colours to see whether we're covering different chemical space. See Fig 5 of the preprint for an example of this kind of analysis by @bendndi .

[MFernflower]

I can confirm dimethylamino and MOM compounds are not in the library it appears there are also no benzyl compounds but a benzoyl is available as EPL-BS0420 and EPL-BS0448

@mattodd @bendndi I assume that F atom is for metabolic stability - not activity at CYP51 - that's more driven by electronic interactions and general molecule shape (see the pdb image I posted)

It would be nice to have a green check next to compounds that can be made with chemicals you already have in lab - yellow box next to stuff you would need to special order chemicals for! @mattodowd25

[mattodowd25]

I could not work out how to do it in DataWarrior...so I used excel. Below is two graphs comparing the EpiChem compounds (blue) and those to be synthesised (orange). One graph compares LogP and molecular weight. Whilst the other compares H-donors vs H-acceptors. There is definitely a lacking of H-donors on both the EpiChem compounds and those proposed to be synthesised. Molecular weight also does not vary too much, with the EpiChem compounds being slightly heavier. The logP does not vary considerably either. With the proposed compounds having on average a lower LogP.

[drc007]

4,4-Difluoropiperidine is not a Hydrogen bond donor?

[fantasy121]

About the p-cyano group possibility of killing potency, I encountered two instances where a change from p-bromo to p-cyano saw a slight increase in potency (compound 3 vs 4 and compound 7 vs 8). However, it did not seem to greatly improve the overall potency (compound 4 vs 8). It seems to me that in these cases, the other parts of the molecules might be playing a bigger role in terms of potency. More from my thesis: https://github.com/OpenSourceMycetoma/Series-1-Fenarimols/blob/master/Theses%20Reports%20Docs/Hung%20Duong%20Hons%20Thesis/Thesis-DUONG-2017.pdf

[MFernflower]

Thanks for the confirmation @fantasy121 !!! A carbonyl spacer is a nogo @mattodd