All Notebooks | Help | Support | About
7th July 2014 @ 13:55

UPDATE 24/7/14

After blogging my first proposal for the synthesis of a fluoroalkene (below), Mat Todd suggested this review (doi: 10.1021/cr200165q) might contain answers to a shorter synthetic route. This indeed was the case, and I have presented a quick revised synthetic route below.


Things to note:

  1. The paper for the first step is doi: 10.1021/jo962373h, the paper for the second step is doi: 10.1016/s0022-1139(99)00172-4.
  2. This paper (doi: 10.1016/S0040-4039(00)79700-0), suggests that Pd(PPh3)4 doesn’t work with heterocycles. The authors used Pd(dppb)Cl2 (dppb = diphenylphosphinebutane) instead which worked very well although the paper is brief and contains no procedures.
  3. I have not yet found a method for synthesising the heterocyclic boronic acid. An alternative could be to substitute it for the bromine on the alkene and couple it directly with the chloro group on the pyrazine or triazolopyrazine. UPDATE: Paper for this using pyridines is here, which I would like to try on pyrizines. 

  4. This synthesis is not stereoselective like the previously proposed synthetic route. However, it would be interesting to test both E and Z isomers so this isn’t too much of a problem (assuming the isomers can be resolved).
  5. The fluoroalkene could potentially be coupled both before and after the formation of the triazolopyrazine core. I will attempt both.
  6. The fluoroalkene is one carbon shorter than in the previous synthesis. This matches up better with the amides and by changing the aldehyde used in the first step I can make fluoroalkenes with 3-carbon chains. 



As part of my coursework for honours I was required to give a short presentation on the Shapiro reaction. During my research I happened across this paper (DOI: 10.1021/ol401637n), which describes the synthesis of fluoroalkenes using the Shapiro reaction. It also mentions that they are isopolar and isoteric mimics of amides, often used in peptidomimetics.  Fluoroalkenes were first introduced as a possible replacement for amines in synthetic peptides in 1986 (DOI: 10.1016/s0040-4020(01)87627-4). Since then they have been well studied in peptides and suggested as an amide replacement in other pharmaceuticals (DOI: 10.1039/b701559c). They are thought to participate in hydrogen bonding (although weaker than an amide) and are dipolar similar to an amide, making it a close mimic.

This had me thinking about how to include them in my triazolopyrazines as amide replacements, not only as an interesting group to test for antimalarial activity, but also for some interesting chemistry to try for my honours project. A representative amide from Series 4 and the replacement fluoroalkene can be seen below, along with an ether from the series for comparison. 

Below is a proposed scheme for the synthesis of a fluoroalkene. This functionality could theoretically be introduced either before or after the triazolopyrazine is synthesised. I will try introducing it first, onto dichloropyrazine for simplicities sake. If this is successful I will also try to introduce it after the formation of the triazolopyrazine core, similar to the synthesis of the ether series.


Feedback on this synthetic route is most welcome. Suggestions for aldehydes that I could incorporate into step 4, and also into the triazolopyrazine core would also be very useful. A downside for our molecules is that the fluroalkene is more lipophilic than an amide, however this could be counteracted by introducing more hydrophilic aldehydes such as pyridine carboxaldehyde or even formylbenzonitrile as shown above.

Papers from Scheme:

1. doi: 10.1021/jo01103a013

2. doi: 10.1021/ja00252a029

3. doi: 10.1016/S0040-4020(01)00076-X

4/5. doi: 10.1021/ic400851w

6/7. doi: 10.1021/ol401637n

Attached Files
9th April 2014 @ 03:40

GitHub Issue #174

The team are currently in the middle of synthesising the triazolopyrazines, both ether and amide linked compounds in the series. I (Jo Ubels) have been looking at the data from MMV on the ether linked series (example below).

I have synthesised the core triazolopyrazine rings (shown in black) with the benzonitrile group attached (pink) and am synthesising more with a chloro group in place of the nitrile. I have made one complete compound with phenethyl alcohol as the ether to test the reaction conditions. My question is, what alcohols (blue), amines or thiols should we buy to link to my compounds? Looking through the MMV data I have found that most of the alcohols already used are fluorinated so I searched for alcohols/amines/thiols with other ring substituents. A selection of alcohols that I have found can be seen in this table along with prices and shipping information.

Table of possible alcohols to buy.pdf

The alcohols that have already been tested can be found in this table, along with their potency data.

Alcohols-Amines that have been tested.pdf

Which compounds should we synthesise next based on the available alcohols? Are any of the possible alcohols particularly attractive/unattractive when paired with the nitrile or chloro group containing triazolopryazine ring cores? We need to be able to reatin potency and increase the solubility and metabolic stability of the compounds. If you can help us decide what to buy and synthesise next then we'd love to hear from you. The list of alcohols/amines/thiols is not exhaustive and I will try to add more suggestions in the next day or so.

To make suggestions you can post a comment below this blog, on GitHub, our G+ site, Twitter or last (and also least) by email to


Jo & the OSM Team

Linked Posts
Attached Files
7th March 2014 @ 01:35

A search in Scifinder was conducted for reaction conditions for the below generic schemes. The purpose of this was to find conditions to conduct a model experiment to create the ether linkage on 2,6-dichloropyrazine before attempting the same reaction with the triazolopyrazine, which we only have in very low yields.

A search for each scheme was performed using the reaction structure search and ticking the ‘substructure’ box. The reactions were filtered by:

  • Yields of 70-100%
  • Journals in English
  • Must have an available experimental procedure.

Promising experiments/conditions for pyrazines:

1. 10.1021/np200386c

Yield of 94%

2. 10.1021/jm061247v

Yield of 83%

Promising experiments/conditions for pyridines:

3. 10.1021/jo800866w

There are several options for the alcohol here, each with different yields.

4. 10.1021/jo901707x


90% where Ar = 2-MeO-C6H4 and R = H

92% where Ar = 4-MeO-C6H4 and R = H

96% where Ar = C6H5 and R = Me

75% where Ar = 4-CF3-C6H4 and R = H

93% where Ar = 4-Cl-C6H4 and R = H


5. 10.1021/jo0602773

Yield of 95%

I will start by trying the reaction from reference 4, but using 2,6-dichloropyrazine instead of the 2-chloropyradine as it is closer to the triazolopyrazine. I will also use phenethyl alcohol (HO(CH2)2C6H5) as this is closest to the ether I will eventually use to link to the triazolopyrazine. I will also attempt this reaction using 2,6-dichloropyrazine and the 4-MeO-C6H4 alcohol listed above as a comparison. Next week I will try the reactions from references 1 and 2 as their reaction times are longer than what I can achieve this afternoon.

Attached Files
30th January 2014 @ 05:02


This week I am starting my honours project at USyd with Mat Todd and Alice Williamson. I’ll be working on the OSM project, specifically focussing on the Series 4 Triazolopyrazines.

Alice has been looking into synthesising the trifluoromethyl form of the below compound since the difluoro form is particularly difficult to make.

My first task was to search PubChem and ChEMBL for similar compounds to see if anyone else had made any. I plugged MMV669844 into PubChem first, adjusting the similarity down to >80% before obtaining any results. 127 compounds were found (example here), and other than two analogs of the above structure (added by Chris Southan for ease of comparison), they all looked rather different.

I then searched for MMV669844 in ChEMBL, this time adjusting the similarity down to >70% before finding any similar structures. The compound with the highest similarity (79.77% similarity) can be seen here.

This compound was also quite dissimilar, and a scan of the rest of the compounds in both PubChem and ChEMBL didn’t yield any compounds with any greater similarity to MMV669844. Series 4 are therefore quite a unique set of molecules as there are no known compounds at 80% similarity or above. Note that this corroborates with a search that Chris Southan performed back in December 2013 detailed on his blog here.


Attached Files