Abridged Transcript of OSDD Open Meeting (17.12.12)

02 Jan
Published by Alice Williamson



+OSDD Malaria, @OSDDMalaria


Matthew Houghton Todd (MHT), University of Sydney, +Matthew Todd, @MatToddChem

Paul Willis (PW), Medicines for Malaria Venture, http://www.mmv.org/about-us/our-team/paul-willis

Murray N Robertson (MNR), University of Sydney, +Murray Robertson, @murrayfold

Alice E Williamson (AEW), University of Sydney, +Alice Williamson, @all_isee

Jimmy R Cronshaw (JRC), University of Sydney, + Jimmy Cronshaw, @JimCronshaw

Kat Badiola (KB), University of Sydney, + Katrina Badiola, @CaverKat

Angela Butera (AB), University of Sydney

Chris Southan (CS), ChrisDS Consulting Limited, +Christopher Southan, @cdsouthan

Ginger Taylor (GT), The Synaptic Leap

Girinath G Pillai (GGP), University of Florida, +Girinath G. Pillai, @giribio

Hong Shen (HS), Roche

Jen Wike (JW), opensource.com, +Jen Wike, @JenWike

Mike Gretes (MG), Oregon State University, +Mike Gretes

Mike Pollastri (MP), Northeastern University @NUTrypKiller

Patrick Thomson, University of Edinburgh, +Patrick Thomson, @patrickthomson


Sanjay Batra, Central Drug Research Institute, +Sanjay Batra

Matin Dean (MD), University of Sydney, +Matin Dean

Minutes kept by:

Welcome and meeting outline:

0 mins

MHT Greets meeting participants and highlights the pre-posted agenda

1. Summary of current aryl pyrrole status and biological results

4 mins
MHT First Round: Aryl pyrroles (OSM-S-5 and OSM-S-6) from the original GSK data set were synthesised and biologically evaluated, confirming activity in whole cell assay. Analogues of the original compounds, plus fragments (to explore the validity of the ester pro-drug hypothesis) and ‘near-neighbours’ synthesised. Modification of the aryl pyrroles led to decreased activity, whereas modification of ‘near-neighbours’ led to increased activity but even poorer solubility.

Second Round: Simple ester replacements (e.g. amide/amine) performed and showed decreased activity. Favourite ‘near-neighbours’ plus OSM-S-5 and OSM-S-6 assessed in further assays (metabolic, HERG, in vivo, late stage gametocyte): a) original GSK compounds bioavailable but readily metabolised (probably due to ester), show low oral efficacy and no activity in late stage gametocyte assay; b) ‘Near-neighbours’ metabolically stable and potent in late stage gametocyte assay (rare and promising) but show very poor solubility limiting bioavailability.

Third Round: Aimed to change ester linker and head group of the aryl pyrrole series. Ester seems essential for activity; small changes in structure destroy activity. Not many changes made to the aryl pyrrole core to date.

2. Recent summary of the biological data: see here and here

8 mins

MHT We have about 100 compounds that have been tested to date. AEW and MNR have highlighted some important trends in the data in order to make decisions regarding the next compounds to be targeted. For example, pyrrole to pyrazole decreased activity, ester to oxazole decreased activity and solubility, replacing ester with amides and amines decreased activity.

3. Where things stand on remaining compounds needed: see here

9 mins

MHT Last remaining compounds (ether, sulphonamide, oxadiazole) still wanted. Sydney team have tried a lot of methods for the synthesis of the ether - all unsuccessful. Still want compounds for completeness and publication purposes.

4. Analyses of the data already posted elsewhere:

10 mins

MHT AEW and MNR performed search of GSK database and looked at other active compounds containing an aryl pyrrole. They then used these compounds to design 5 hybrid compounds that they are keen to synthesise and send for testing along with 5 more speculative compounds. see here

5. Questions arising:

11 mins

MHT Outlined three important questions to be answered in this meeting:

• Do we continue with some more compounds of aryl pyrrole series?

• Should we revisit the ‘near-neighbours’?

• What does modelling data (MNR) suggest?

12 mins

PW Is happy with the introduction to the meeting and keen to discuss the questions outlined.


• Do we continue with some more compounds of aryl pyrrole series?

13 mins

MHT We still want to synthesise the missing three compounds (ether, sulphonamide, oxadiazole). Ether proving to be a real pain, although we have a quote from a CRO willing to make this compound on a no-win, no-fee basis. MD working on the sulfonamide, PT has also volunteered to have a go at this compound, Sydney team happy to help in anyway (send material, preps etc.) but would aim to get the compound tested locally (UK or Europe) rather than sending to Australia.

14 mins

MHT As PW has previously said, we don’t want to spend too much time making compounds that might be inactive. However, based on active compounds from the TCAMS set, Sydney team keen to make five representative examples from the aryl pyrrole set that seem promising and look to be synthetically accessible.

17 mins

MHT Asks PW what he thinks about the Sydney team spending another month working on the aryl pyrroles and in particular the hybrid compounds.

PW Asks which of the hybrid compounds we want to make.

AEW Says that the Sydney team want to make the first 4 compounds drawn in the first targets box’‘. The first two structures represent fragments of OSM-S-6. MNR pointed out the symmetry in this compound, so it would be interesting to test this fragment. The second two compounds are hybrids of OSM-S-5 and two active compounds from the TCAMS set.

21 mins

PW Thinks the two esters are risky as the ester needs to go in order to obtain an orally active compound). Could be a tough/long journey as if active we will then have to replace the ester – not such a good idea at this stage of the project, better to start a new series. Thinks the fragments are interesting and asks AEW and MNR if they were in the original HTS. Asks AEW and MNR to search back through ChEMBL to see if they are known. 5th compound interesting isostere but looks challenging (chiral centre) and so not worth a long programme of synthesis. In summary, if fragments are commercial or quick to make then go for it, forget about esters and the 5th compound is highly speculative so only worthwhile if quick to synthesise.

22 mins

MHT Thinks the 5th compound could be accessed from the aldehyde. Attempts to replace the ester haven’t been successful. Looking at the metabolic stability (Sue Charman), the esters were poor in mouse plasma but good in human plasma – are we unfairly critisising the compound for being less stable in mouse models?

PW Says we could make a case about increased stability in human plasma but mouse models are integral part of screening in the drug discovery programme and so compounds need to be stable in mouse plasma in order to take a compound forward as safety tests are performed in rodents.

25 mins

MHT Going back to the ester, we have made a number of changes including exploring the ‘pro-drugs’ but all of the compounds have been less active than the original hit. MNR has done some preliminary work modelling the active compounds to try and find out what properties are important for activity. His model suggests that the ester is important for binding.

27 mins

MNR modelling might not be so useful as results suggest that the ester is important In fact, everything points towards the ester. Substructure search of ChEMBl database gives 40 aryl pyrrole containing compounds that can be clustered into 5 sets. Apart from ‘near-neighbours’ all contain esters. (Red spheres are hydrogen bond acceptors, green spheres are hydrophobic regions, orange circles are aromatic pi-stacking areas.) The cluster of compounds all map, including ‘near-neighbours’. Modeling completed prior to receiving third set of biological data. Prior to receiving the data, OSM-S-103 including carbonyl but not ester was predicted to be active, however it was only poorly active. Data so far suggests that ester linkage required, needs another hydrogen bond acceptor, examples synthesised with nitrogen or hydrogen bond donor in this position are inactive. Therefore, we need an ester isostere. Could use this modeling hypothesis when looking for a new core/series for the same target.

Appeal for help on Twitter by more experienced modeling teams/people with more software power seem promising – their expertise could be invaluable.

• Should we revisit the ‘near-neighbours’?

• What does modelling data (MNR) suggest?

31 mins

MHT Interested in the modelling results, particularly that the ‘near-neighbours’ fit the model despite the lack of the ester.

32 mins

MNR That’s right, one carbonyl hits one of the features very nicely and the sulfur on the ring appears to be acting as a hydrogen bond acceptor. Thinks that we need to get away from ester towards an active bioisostere.

MHT Should we be looking at the 'near-neighbours’ (poor solubility and red flag of Michael acceptor)? Paul Ylioya tried to perform Michael additions with thiol and reduction with hydride and found it to be very unreactive. A more realistic glutathione trapping experiment on one of the ‘near-neighbours’ proposed by Sue Charman. MHT asks PW if he is solidly against the ‘near-neighbours’.

34 mins

PW Thinks ligand efficiency is problematic. ‘Near-neighbours’ show similar potency but have much larger molecular weight, therefore lower binding efficiency – heading in the wrong direction. We need to harness these interactions but with smaller molecules that bind efficiently.

35 mins

MHT OSM-S-35 discussed (26 nM). A small molecule that could be tinkered with. The double bond is unreactive and not acting as a Michael acceptor. Near neighbours are also perform well in the late stage gametocyte assay whereas the original GSK compound (OSM-S-5) doesn’t. Questions if this is a priority? State that this is interesting as it highlights the distinguishing feature between the two sets.

37 mins

PW It’s nice to have but tied into that we’ve got to get oral activity, as this is the key feature we are lacking at the moment. Perhaps worth having a think of related structures we could make that are more drug like, a small set of targets.

MHT Suggests that some of the near neighbours can be sent to Sue Charman for trapping analysis.

39 mins

MHT Questions the core of the near neighbour and if Paul would prefer us to stay away from it. PW It’s not a definite no. Also interested in using the pharmacophore model to predict the activities of the most wanted compounds. If they are not likely to be active then they could be parked for the moment.

MNR Has mainly focused on the ether compound. Couldn’t get the ether to fit and doesn’t think the ether should be any more active than OSM-S-103. Not worked on the other 2 compounds yet.

MHT Asks if the modeling could be used to predict target.

MNR In theory you could screen compounds with kinase panel data and look for trends

42 mins

MHT Suggests looking for compounds that have DHODH (dihydroorotate dehydrogenase) activity and trying to map them.

44 mins

PW States that this is an interesting exercise, but in the drug discovery strategy it doesn’t move things forward for the early lead criteria. This could just be run on the side as an interesting activity

MHT Was thinking more towards a paper, but agrees that it doesn’t help with the lead discovery process.

45 mins

MHT Summarises the options if we are moving away from the aryl pyrrole series.

- Tweaking the near neighbours.

- Run with one of the compounds on the back burner

- Triazolourea (OSM-S-56) but we understand there may be another group interested in that?

- The thienopyrimidine series as the original hit has now been synthesised but it’s still to be tested.

A number of these are going to be made by Angela, summer student, in the Sydney. If not these two then we have to think about something else. But before we embark on a new series, we’d like to discuss with GSK and MMV to make sure that we are going after the right compound.

47 mins

PW Has already discussed with GSK what is already being explored and suggests it best to suggest series from the paper then he can advise on whether they are already being followed up or not. He states than another possibility is the malaria box. Select one of those and try to improve its potency. He also suggests that it might be worth a topic for the next meeting just to put on the table the options that we might want to consider for a new series then try to select one.

49 mins

MHT Next meeting next February? As this will allow time for a little more synthesis and evaluation.

PWThinks we need to prioritise what is going to be done in January.

MHT Agrees with PW, esters are a weakness. Would prefer at the moment to explore the 'near-neighbours'. At the same time, validate and explore the thienopyrimidine. Also, the sulphonamide that PT is going to try.

PW We should still use the pharmacophore model to prioritise the most wanted compounds. If they don’t look promising and the synthesis is tough then there is not a strong case to continue and make them.

53 mins
PW Agrees that the suggested plan looks good.

54 mins

MHT Welcomes any new synthetic chemists to join in and make suggestions. Will plan next meeting early and give about 1 months’ notice. Is middle of February OK?

PW Agrees with mid-February

56 mins

CS Has recently posted a blog post regarding target mapping and method of action. GSK have reran the TCAMS set and have published a lead.

MHT We’ve already discussed with GSK and they’ve agreed to run the DHODH but this needs to be chased up.

CS Offers to do work to identify assays and people working on these. Asks Paul if target identification is crucial?

60 mins

PW States that target identification is not crucial for MMV as they can take things forward into man without knowing the target. Obviously it is a bonus but not essential. It’s worth doing the exercise but finding people still working with the assays may be a problem. If, with using in silico predictions, we can focus to a particular target rather than screening against everything, then this could be useful.

CS States the advantages of knowing the target of action. HTS, etc.

65 mins

MNR Asks about the potential of purchasing a small selection of commercial compounds based on pharmacophore results. He can easily generate a shortlist to post online for discussion.

PW Agrees with the idea of an online shortlist and discussion.

MHT Thanks attendees and closes meeting.