Sandbox for Synthetic Ideas

Published by MatTodd on 30 January 2011 - 7:49am

This is a post of random synthesis ideas for praziquantel that have been sitting in my inbox for too long. Both sources are industrial. If anyone has any comments, please post. If anyone has any further ideas, please post directly here – you can see how long it can take things to emerge if you use email...
 
Back in June I gave a talk at Stanford to a group called SPARK hosted by Kevin Grimes. This was a very interesting collection of people - highly interdisciplinary. Students and staff give talks, and the idea of the meetings is to brainstorm ideas on the projects underway, many of which were relevant to tropical diseases. These were like normal science group meetings, but with a much broader audience who were not shy of asking tough questions on detail and overall project direction. I really liked the way that a) students were actively involved along with faculty in meetings, and b) much of the stuff under discussion was in progress and incomplete, allowing real input from the audience.
 
My talk about the stuff The Synaptic Leap's project with praziquantel was interesting because of the diverse nature of the questions I got. This is one of the benefits of talking to a mixed audience of smart people. One of the most vocal skeptics of the approach was Steve Schow from Telik. His questions were excellent because they were grounded in economics. Essentially the arguments centered around where the money would come from for open compound development (or even drug discovery). I won't rehash the discussion here, because that's for a separate post that's building, but after the meeting Steve (who is an organic chemist by training) sent in a bunch of suggestions for other syntheses of enantioenriched PZQ. After checking with him, I am posting them here - his schemes posted directly. I'm posting them for completeness, in case the ideas chime with anyone or anyone wants to chip in. I want to thank Steve for contributing them.
 
This first idea is essentially the approach that's already been described. DSM are currently conducting catalyst screens for this reaction on the enamide shown and Development Chemicals had a hit with a related benzoyl substrate.
 

Steve's appeal to literature was the following:
 



Most of these examples have coordinating groups such as carbonyls that are well-placed to deliver the metal (and H) from a defined side. This has already been discussed here. There are a couple of examples above that do not require this. Perhaps the relevant ligands would be of interest - if anyone has the literature references for these? Steve's second suggestion involves an asymmetric hydrogenation at an earlier stage.

With the following precedents:

We have found working with dihydroisoquinolines to be rather annoying, given how prone they are to oxidation, but certainly the chances of a successful asymmetric hydrogenation here look to be better. Anyone have any comments on this approach, or literature sources for a comparison? Steve's final suggestions were as follows:

We'd thought of the first two, but have not done much/any work on them. It's a challenge because the enamide is not terribly reactive - it takes a lot to protonate the molecule. In the middle case, the tautomer shown is less stable than that where the double bond is conjugated with the ring. In the final case, the asymmetric reduction approach is no use since introduction of acid is required for the cyclization and this produces the acyliminium ion which scrambles stereochemistry. The introduction of the chiral acid is essentially the problem we're having with the Pictet-Spengler approach, in that the Ar ring needs methoxy groups to be active in this reaction. Thus far we have drawn blanks.
 
More recently, the guys at Creative Chemistry, who were looking into a route to enantioenriched PZQ, and developed a resolution approach, also contributed a few ideas they had been playing around with, but had not pursued. Again, I want to thank them for sending these along, which is in the spirit of this open project. Apologies for taking too long to post these. The following is essentially as received from DC. Idea 1: Reaction of tetrahydroisoquinoline with peroxide/tungstate/cyanide gives the unsaturated nitrile (TL, 1987, 6469). Yields when this was attempted were lower than described and the product was obtained by distillation. It may be possible to reduce this material enantioselectively, but the centre adjacent to the nitrile will be very acidic. This could also offer an easy entry to the diamine for resolution.

Idea 2: The asymmetric synthesis of praziquantel via a Bischler-Napieralski reaction was attempted as described by Czarnocki. Unfortunately, on reading the experimental, it became apparent that the 96% yield of the B-N reaction was based on recovered starting material and that the yield of product was only 15%. In our hands, the reaction gave several products and very little remaining starting material. Using the simpler chloromethyl amide the reaction went very cleanly. Unfortunately, the material is possibly too unstable for any further reaction unless kept under acidic conditions. 

Idea 3: The preparation and hydrogenation of the tartaric acid derivative has been documented (Can J Chem, 1992, 70, 1555 (MHT - annoying journal website - the paper is meant to be here); Acta Chim Acad Sci Hung, 1976, 89, 161 (seemingly not online)). It may be possible to perform this reaction with the unsubstituted analogue shown. If so the amine could be reacted with glycine followed by cleavage and reductive amination (or cleavage/reduction to create a chirally stable centre). A similar reaction using R-(+)-glyceraldehyde has been documented and, after hydrogenation, gives product of the desired stereochemistry (Can J Chem, 1986, 2205).

Idea 4: Speculatively, a suitable B-N product could be converted to the amide and the resulting product hydrogenated using, for example, a Binap type catalyst. This type of reaction has been documented.

Comments

 Matt
Thanks for putting the ideas up for people to have a look at. 
 
We did investigate the unstaurated nitrile reduction further as a means to the racemic diamine and those results have been posted here.
 
As for all of the chiral reduction approaches, we still firmly believe that the price of the metal (especially iridium and rhodium) is going to make this non-viable on such a low added value product and that resolution/recycle is the only economically viable way forward (this is an incredible belief given my colleague's 18 years in chiral reduction methods!).
For instance, in our work with mefloquine, we were able to perform a chiral reduction of the ketone intermediate using a Ru transfer hydrogenation method. With a TON of 2,000 this gave a cost contribution to the finished product of about $20 which was just acceptable given a target figure for the required cost of production which was several times that of PZQ (mefloquine sells for 10x the price of PZQ).
 
Also, once again, this is all from Creative Chemistry not Development Chemicals; they are involved in coordinating the projects from our research and getting development partners involved for the scale-up.
 
Bill
 

MatTodd's picture

Sorry, Bill - attribution now corrected above. Yes, I also note your comments on the impact of cost, and the difficulty of finding a catalyst. At this stage, we're looking around to see what kinds of catalysts might be able to effect the transformation, so the project has a slightly academic flavour, because the reaction is so difficult. It's interesting to see whether anyone can be predictive about what might work for this substrate - to date we've been relying on screening. But this is certainly a sub-branch of the overall project, for the perfectly correct reasons you lay out. So in this case we worry later about the cost, depending on what we find.

As for idea 2, it has been reported in the patent literature that related chloromethyl compounds have been reacted with secondary amines in pretty good yields and then reduced catalytically and so there may be some possibilities here [see EP0330360].