Devise Synthesis for 24 and 25

[alintheopen]

We need a synthetic route to these two molecules. If you have time to do a lit search then please be our guest?

[PatrickThomson]

I'm doing this now and will post my findings here momentarily.

[PatrickThomson]

The boronic acid needed for 25 has been synthesised before, in a 2010 BMCL: http://www.sciencedirect.com/science/article/pii/S0960894X10014587 - from the bromo-methyl-sulfonamide which is commercially "available" at >$1000/ fopr 5g (CAS 1032507-35-9). The bromide is made from the bromo-methyl-sulfonylchloride (886501-61-7) which is available from a few suppliers for $250/g - this compound is a key intermediate.

The key intermediate is fairly poorly-known in the literature, being made either from the dibromotoluene with benzylthiol then SO2Cl2 (patent WO2009/103440 A1, 2009), or the corresponding aniline via a sandmeyer then sulfur dioxide (med chem lett, http://pubs.acs.org/doi/abs/10.1021/ml2000627). I favour the former route, although it has 2 steps and uses palladium catalysis with a reasonably fancy ligand, the second step to convert it to the acid chloride can be telescoped with the ammonia quench and overall the synthesis would I think be shorter and easier, and avoid the use of liquid SO2 and diazonium salts. Given my limited budget (£700 until forever) I think I'd be better off synthesising the key intermediate (and borrowing the palladium) rather than buying it, as it only adds one full reaction's worth of time to the workload. Additionally, the suppliers aren't the most well-known and moisture-sensitive commercial material always has the chance to turn up half-hydrolysed.

[PatrickThomson]

The sulfonic acid could be trivially converted to the sulfonyl chloride; there is a literature example where 2-bromotoluene is sulfonated but it's 120 years old and in German.

[PatrickThomson]

Nothing useful with an iodine-for-bromine switch either.

[PatrickThomson]

The 4-chloro-7-methylthiopyrimidine that I think is the correct analogue, for 24, is commercially available for £34/gram and I'll just order it; tweeting to Alice just to check that I have the right synthesis in mind before I order it:

[mattodd]

Yes, that's a nice precursor for 24 - saves some time! The sulfonation of 2-bromotoluene also seems to be the way to go - if you can get the paper we must be able to get it translated.

We'll need a plan for how we'll do the Suzukis, or at least how to get the two halves of these molecules in the same country. It's might be an option for us to buy chemicals and get them delivered overseas, but there may be a small legal problem there I'd need to evaluate. Do you have access to a microwave synthesiser, Patrick?

[PatrickThomson]

Ok, I've ordered the precursor for 24, and I think I can still make the case for the dibromotoluene route to the boronate as follows: It's a synthesis and a separation, using borrowed materials, followed by a simple telescoped benzyl->chloride->sulfonamide that'll take a day with no silica, 2 workups and reasonably modern documentation throughout.

The 2-bromotoluene route is actually in english here: http://pubs.rsc.org/en/content/articlelanding/1892/ct/ct8926101023 but would involve speculative research due to the scarcity of documentation - and the fact that such a simple transformation hasn't been reported after 1900 makes me think that something funky happens instead. Even if it worked, I'd be saving at best 1 column and no time, followed by an equivalently-simple acid->chloride->sulfonamide. I could spend the same length of time trying to separate a mixture of ortho- and para-sulfonic acids as it would to synthesise the sulfonic acid from dibromotoluene.

I have several microwave reactors I have access to, and I think the way to go with this route is that I'm sent samples of both halves of OSM-S-106 - the parent aminothienopyrimidine (saves repeat work etc). From what I can tell, it's a lot easier to export research samples from AUS than it is to import them. Then, I make all 4 permutatons of methylated and non-methylated (possibly the bis-methylated would be a mix of atropisomers but we'll cross that bridge when we come to it ...). This means we also get to "calibrate" the screens at dundee with a copy of OSM-S-106 that I'd make myself. This calibration might be useful if these compounds are tickling the lower limits of solubility as the assays may respond differently to aggregates etc.

[alintheopen]

Hi Patrick, I like your plan. I think it would make sense to start with SNAr on the commercial 4-chloro-7-methylthiopyrimidine and then the bromination. We could then send you samples of the boronic esters below (once we've made them all) for you to couple to make derivatives of type 5. If you also have a go at 6 and maybe 7 then we could evaluate where Sydney and Edinburgh teams are up to and then either you can send us the boronic esters or we could send you the relevant bromides... We could just send you some OSM-S-106 to calibrate the Dundee assay rather than you resynthesising. How does this sound @mattodd and @PatrickThomson? Cheers Alice

[PatrickThomson]

Sounds good - I don't know what the lead time is like for getting samples out of Australia, so design and synthesis of 7 will give me a backup plan if shipping takes forever. I only have a gram of 5 (£38), so I'll look at the higher value targets for 5 R1 and R2 for maximum data overlap with the des-methyl series.

[mattodd]

Sounds good - you only have a gram of which, sorry, Patrick? Lead time - well the shipping is usually just a few days - we've had good experience with this to date. Not sure about completion of synthesis of those boronic esters. Seems like a good split of effort, though.

[PatrickThomson]

Ok, here's my plan: Crack on with the synthesis of 2 and 5. If I get to the end of both of those synthetic routes before samples of 10 and 11 arrive, I'll cross them with each other to make 7. If obtaining samples of 10 and 11 is going to be difficult, I can resynthesise them but I'd rather keep on novel compounds. I probably only need ca. 100mg of each, but as much as can be spared will be very useful! 3 is a great suggestion which I'll research now and switch over to instead of 2 if it's vastly substantially easier to make but the 1-step precursors cost the same ludicrous amount, so probably it won't be easier.

[PatrickThomson]

Good news! 3 can be made in 2 steps by brominating p-toluenesulfonamide, then converting to the bpin in the usual way. Bad news! The first step uses bromine as a solvent. It's probably still worth doing it first/instead of 2 though. Mat, has there been modelling work done on 2 that would be wasted if I switched to 3?

I'm also going to source some chiral chromatography expertise here in edinburgh incase 9a and 9b don't interconvert (and other compounds as appropriate).

[mattodd]

You mean 8 where the methyl is switched over? No, no modelling AFAIK - the assumption is that ortho to the thiophene will buckle it.

[PatrickThomson]

Yeah, I could make both just to see what the results are but I'll probably prioritise one of them based on whatever's in stock.

[alintheopen]

Looks good Patrick, I will assess where we're up to with boronic esters next week and we can have another chat. I think 11 is pretty scarce so we would probably have to resynthesise but I will confirm. As far as 3 vs 2, I think that we might need both. The idea being (at least in my mind) that the clash between H or Me group on the thiophene and the bulkier group on the arylsulfonamide would cause the aryl sulfonamide to orient itself in opposite directions. I have pasted a cartoon below to demonstrate my thinking.