macrolactonization vs. ring-closing metathesis

[Dan_Ny]

To get large rings (with a endo lactone functionality) closed, the traditional approach has been to use macrolactonization. Because the entropy factor overrides the enthalpy factor for large rings, suppression of the intermolecular reaction becomes more difficult. That problem was addressed by activation of the carboxylic or alcohol moiety, for example by the use of thioesters to "push" the reaction against entropy by increasing the enthalpy (correct my, if i am wrong here).

in the last time, i came over an immense number of examples where rings were closed by ring-closing metathesis (thanks be to grubbs). is the time of macrolactinzation finally over?

 

opinions, please!

[Horza2002]

The compound that my PhD is based on poses 10- and 12-membered carbocylic rings. In every total synthesis reported for these compounds, the ring is formed by ring closing metathesis (RCM). This method is far more efficient at forming larger membered rings than "traditional" methods, although it isn't without its problems. The size of the ring, while not as restrcited as traditional methods, still affects how efficient this reaction is. Steric factors, especially with groups bound to the alkene, also have an issue.

 

Another slight problem, is that Grubbs catalysts catalyse a whole range of reactions, not just ring closing metathesis. It is not underheard of to get a competing ring opening reactions and polymerisation reactions. With all that said though, metathesis is a very powerful synthetic tool.

 

Biology, in macrocylce chemistry, is certinaly superior to anything that we chemists can currently do. There are a massive range of macrolids...the most well studied of these are the Erythromycin's. This is a PKS (polyketide synthase) natural product that conains a macrocyle that is formed when the completed polyketide chain is transfered to a thioesterase domain and cyclised. This is a very common step in biological cyclisation reactions.

 

Whereas Erthromycin's cylisses onto an ester, more often it is onto a thioester...hence the development of the thioester cyclisation methods. Using 2,2'-Dipyridyl sulphide and triphenylphosphine is modelled on how biology does these reactions. I have a lot of experience using (PyS)₂ reaction, however im using it in a different reaction and not a macrocyclisation.

 

 

In addition to this, I think this example shows how amazing bioloigcal systems are at synthesising compounds...Erythromycin's are produced in almost complete enantioselective. Very impressive!!!

Edited March 19, 2011 by Horza2002

[Dan_Ny]

Wow. I am beginning to like this stuff. Hm, so Grubbs also has it's disadvantages. And enzymes - I definitely don't know enough about biocatalysis. Are enzymes allowed to be used in total synthesis? Or ist it "too biological"? The problem might be that you don't know if a specific enzyme also works for your substrate....

By the way, do you know how they get regioselectivity for the sugars in erythromycin D?

Edited March 20, 2011 by Dan_Ny

[Horza2002]

The same as any catalysis really, there are a number of disadvantages. The main one is that are the two ends of the double bonds that you want to cyclise get further and further away from each other (i.e. to make larger and larger rings) the efficieny of the reaction is lost. Heres a review that talks about it. Don't get me wrong, methathesis is a very powerful tool....you just need to give several subtle tweaks to your individual reaction to get it too work (i.e. concetration, use of catalysts, use reduced pressue, etc.)

 

Schuster, M. and Blechert, S. (1997), Olefin Metathesis in Organic Chemistry. Angewandte Chemie International Edition in English, 36: 20362056. doi: 10.1002/anie.199720361

 

Enzymes can be used in total synthesis, or in general synthesis yes. The problem is, however, the majoritiy of enzymes are specific for one substrate and only one substrate. However, there are a lot of enzymes that can be used. One example I have used is an acetylation enzyme. Basically I made a non-natrual amino acid racemically and then treated it with this enzyme. This enzyme then only reacted with the L-form of my racemate to convert it to the N-acetyl functional group. At the end of the reaction, I simply did an acid wash to remove the D-form and then hydrolyised the amide again to giev me nice enatiomerically pure L-amino acid. This type of resolution can be useful, but it often only works on very small molecules; normally ones that resemble amino acids. You can also used a compination of this enzyme and a racemise enzyme for amino acid like molecules. The racemise enzyme racemises the remaining D-amino acid to the L which is then reacted so you actually get a 100% yeild of enatiomerically pure compound and not just the 50% you put in. Heres a review on the matter:

 

Enzymes for chemical synthesis, Kathryn M. Koeller & Chi-Huey Wong, Nature, 409, 232-240, 2001

 

I'd like to highlight the enzyme aldolase from that review. Using that enzyme, it is possible to get enatioselective forming of carbon-carbon bond via an aldol reaction. Very cool!

 

The regioselectivity of the the sugar moeities will arise from the enzyme specificitiy when it catalsyses the reaction. There are other types of Ethromycin compunds (A,B, C, D and I beleive E). These basically differ from the number of suagrs, the regiochemistry and some further modifications to the macrolid ring.

Edited March 20, 2011 by Horza2002

[Dan_Ny]

I just read this review about cross metathesis, unfortunately it did not cover too much RCM...

Stephen J. Connon and Siegfried Blechert*, Angew. Chem. Int. Ed. 2003, 42, 1900 – 1923

I'll read yours now. You can't know enough about this. There are some interesting feautures in cross metathesis, for example chelation of the grubbs catalyst, low reactivity of electron-poor olfins, ROMP-CM cascades... cool stuff.

Ah and thanks for the review about enzymes in synthesis.

Edited March 20, 2011 by Dan_Ny