Fresh Round of Evaluation vs Aldehyde Oxidase

[mattodd]

We received new microsomal clearance data (#484). We have been kindly offered some fresh analysis vs. Aldehyde Oxidase (AO) by Scott Obach at Pfizer. Two analyses can be done:

1. AO assay. Evaluation to see which compounds are substrates for the enzyme, as per the last such assay we did on Series 4, which showed there was a range of activities.

Now, to my mind, we ultimately want to know the stability of these compounds in the presence of hepatocytes (heps). If compounds are cleared quickly in the in vitro microsomal assay there seems little point in evaluating vs. heps. However, if compounds are relatively stable to microsomes, we should consider evaluation vs AO. If the compounds are found to be stable, we would have solid reasons for expecting good results vs heps. If they are found to be unstable, then we would expect similar results from heps. Naturally this means that we would, either way, want heps data on the compounds (!) but we would be better placed to interpret those data. Would people agree?

Suggestions (structures all shown below).

a) MMV688896 and MMV897708 were found to be particularly stable to microsomes (#484) and should be evaluated vs. AO.

b) In the previous AO assay, the core-changed compound MMV669846 was found to be a reasonable substrate. We could now evaluate frontrunners MMV639565 (same but parent core) and MMV663915 (ditto but lacking Ar fluorines) to compare.

c) A previously evaluated compound was the alcohol MMV670947. We have to hand the analog lacking the aromatic F’s, MMV693155. Worth evaluating, or are we already interrogating the importance of the Ar-F’s in possibility 2 above?

My inclination is to act on a) and b), not c)

2. AO Derivatisation. Scott runs a super nice protocol that uses AO as a synthetic tool. (Lit refs: here and here) Scott is offering to take a mg of a couple of compounds and have AO modify the compounds, then isolate the products ready for potency analysis, to see if the products are themselves active. The ideal substrate is one that is a good substrate for AO, and which is active. We don’t have this. What we do have is as follows:

i) A potent compound that is converted slowly by AO: MMV897709 (racemate of MMV669844)

ii) Well-cleared, but non-potent compounds MMV670246 and MMV670944. We could of course see if the conversion by AO converts the compound into a potent derivative.

My current inclination is to send all three compounds. We have stocks.

Any advice/thoughts on all this?

Zip of Chemdraw file if needed:
AO Planning.cdx.zip

[MedChemProf]

@mattodd My initial reaction was to agree to your suggestion of just using the pairs MMV688896 / MMV897708 and MMV639565 / MMV663915 in the AO assay, but I am a little concerned that the results might be misleading without also looking at MMV693155. With the limited amount of AO activity data we do have, I think I could make the argument that AO oxidizes the triazolopyrazine ring primarily. Furthermore, the pyrazolopyrazine ring is more susceptible to AO oxidation if substituted with an electron withdrawing amide.

I would also like MMV693155 in the AO assay because I think that we are actually looking at multiple metabolic liabilities and it may help to have this extra data point to deconvolute the contributions from AO versus a particular CYP. Additionally, it would be ideal if we could get HLM data on MMV670947 (if material available) to absolutely identify the relative oxidative contributions of the two enzymes. MMV693155 has essentially identical LogP, LogD and solubility as MMV688896, but it is metabolized 7X faster than MMV688896. As discussed previously, the microsome assay largely shows CYP oxidations versus any contribution from AO. Furthermore, the oxidations of MMV639565 and MMV663915 in the HLM already inform us on the contributions of the 3,4-Diflouro’s. Both of these compounds may just be oxidized to the benzylic alcohols similar to MMV688896 and MMV8977908 and continue to be active with lowered potency. (Unfortunately, my argument does fall apart to a large extent when comparing MMV688896 and MMV672687).

Sorry for the rambling discussion. Bottom line, inclusion of MMV693155 (and MMV672687 if material available) might be enough data points to clarify the role of AO. I think it more important to get the AO activity data at this point versus running the AO Derivitization assay.

[alintheopen]

Great news. Previously measured AO activity has been added to the Master Chemical List and to the 'daughter' Frontrunner spreadsheet (cc. #484)
How much sample is required for these assays?

[mattodd]

OK, summary so far:

AO screen: MMV688896, MMV897708, MMV639565, MMV663915, MMV693155, MMV672687
AO conversion: MMV897709, MMV670246, MMV670944, MMV670947 (if have) or MMV693155 (see below for more on this one).
Chase: HLM data desirable on MMV670947 to compare with the non-fluoro compound MMV693155

Note from Scott by email: "From the lead diversification assay (AO used as a synthesis tool), the material that results will be at a concentration of 0.2-2 mM in solvent (d6-DMSO), and there will be only be 35-40 uL of this solution, I want to make sure that such material would suit the needs of whatever your primary assay for potency is. It is a little different thinking on this small scale—one has to be sure to dilute the solvent out to an acceptable level in the assay and that the assay volume is such that you’ll have enough stock solution to make an inhibition curve. In some cases, we can increase the substrate concentration a bit, but with enzymatic conversions one doesn’t get back a proportionate increase in product.” I will ping Irene Hallyburton on this point.

2 days ago Scott also checked in by email, after reading the above:

“For using AO to generate a new lead, I’ll test not only human cytosol but also cytosol from mouse, rabbit, and monkey because it can frequently be the case that even if human AO only does slow turnover, one of these other species’ AO may do great conversion. So don’t let slow human cytosol turnover dissuade you from selecting a compound for this—I always test others before selecting which species to use in the biosynthesis. (But if we did see no turnover at all, then that probably is not a good compound to try.)”

This is awesome, so potent/interesting compounds with slow AO turnover (e.g. MMV897709) are OK for the moment. This means we could also send MMV670947 if we have it, or its non-F analog the frontrunner MMV693155 that we’re sending anyway for the AO assay. Included in the summary list above.

@edwintse @alintheopen we have 2-3 mg of each that we can vial up, right? If so, can you please get these ready?

@MedChemProf if you’re still good to send some OSM-W-7 (#487), I’ll send shipping info by email. An excellent addition in my view.

[MedChemProf]

@mattodd I just weighed out 2.3mg of OSM-W-7 into a vial.

[mattodd]

Clarification from Irene Hallyburton on the relevant conc ranges. Have asked Scott to verify whether this is OK - essentially whether the outputs from the AOX diversification experiment are suitable for blood-stage potency evaluation directly.

"if I understand correctly the concentration will be pretty low (0,2-2mM). For a compound of unknown potency we would normally need a minimum of 10ul of a 10mM DMSO stock to get a concentration range starting from 10uM. This is 0.1% DMSO in the assay. For more potent compounds we can start with a top concentration of 1uM, using a 1mM stock, but we would still require a minimum of 10ul. So, if they could get as close to 2mM as possible, and get 10ul, we could start the dilution curve at 2uM"

Getting ready to ship on the next day or so. Will ping you @MedChemProf