total synthesis of natural

product isolated from

liverwort Radula perrottetii and Radula

variabilis – Radulanin H.

Its structure is shown below:

Radulanin H exhibits inhibitory activity aganist cyclooxygenase. It has no stereocentres but contains seven-membered heterocyclic ring with trisubstituted double bond and highly-substituted aromatic ring. Concise synthesis of this aromatic ring is the challange.

So how this synthesis was planned?

It’s not a surprise that ring-closing metathesis was used in contruction of heterocyclic ring. This simplification reveals compound A which can be prepared from phenol B by introduction of two allyl groups in Claisen and Williamson reactions. Now, question arises – in which way aromatic ring B (with all substituents on their places) can be synthesised? Well, for such complex system authors of paper have chosen synthesis from acyclic substrates:

It’s not very obvious how this can be done by using ethyl acetoacetate and benzyl bromide – so let’s see how this synthesis has been acomplished.

The answer is – dianions! They acted on ethyl acetoacetate 2 with a little bit more than 1 eq NaH and they got of course anion 2a. For this moment it sounds like simple synthesis from acetoacetates. But they didn’t add to reaction mixture  electrophile but another equivalent of strong base instead! In this way they got dianion 2b. When in next step they added benzyl bromide more reactive anion reacted with electrophile and ketoester 3 was formed. That’s not standard procedure of alkylate ethyl acetoacetate Ketoester 3 was in next step protected as acetal 4. Yield after two steps is good.

Well, next step involves also dianion strategy but this time dianion 2b is acylated with previously formed ketoester 4 (that’s why carbonyl group in 3 had to be protected) and we have compound 5.

Ok, now 5 should make some aromatic ring It can be done by deprotection of acetal 5 to polycarbonyl compound 5a which forms 5c by some aldol-like intramolecular reaction. Now – by using weak base sodium perchlorate – 5c is converted to 6. Yield is only 48% but in one step you get highly-substituted aromatic ring.

Now 6 is converted to its allyl ethers mixture 7a and 7b. These compounds aren’t isolated and in high temperature, under Claisen rearrangement reaction conditions, allylbenzene 8 is formed.

Next step is second Williamson reaction and in this way allyl ether 9 is formed. Regioselectivity of this reaction is quite obvious because other OH group between two substituent is of course less reactive. Cyclisation of 9 to 10 probably is not very easy because paper’s authors used 20% mol Grubbs’ catalyst. It’s a quite big amount, but yield after two steps (RCM and deprotection step) was very good.

For more details see: M. Yoshida, K. Nakatani, K. Shishido, Tetrahedron, 2009, 65, 5702–5708.

For Polish version of this post please see here.

Ortho strategy has been chosen to construct such trisubstituted pyridines. 4-substituted pyridine seems to be a good starting material in such strategy. Let’s see strategic disconnection in retrosynthetic plan:

Total synthesis should be quick.

As you can see it starts with 4-cyanopyridine 2. First and second step are of course halogenations steps which go through ortho-lithiations. In first step bromine is introduced into molecule (through nucleophilic attack on CBr₄) and in second step – iodine. LTMP is lithium 2,2,6,6-tetramethylpiperide, a strong and sterically hindered base:

Next steps allow to construct five-membered ring:

Conversion 5 to unsaturated ester 6 is Heck reaction. Unsaturated ester 6 can be then hydrogenated on Adam’s catalyst to 7. 7 in turn is cyclized to 8 under basic conditions (enolisation and nucleophilic attack on CN group) and subsequent hydrolysis.

From ketone 8 both louisianins can be prepared under Stille coupling reaction conditions.

The only difference between those two reactions is adition of base (DBU = 1,8-diazabicycloundec-7-ene) which causes isomerisation to more stable (trans and disubstituted) alkene.

That’s all. Happy New Year and see:

Chang, C. -Y. et al., Tetrahedron (2009), doi:10.1016/j.tet.2008.11.075

Let me add that aplysamine 6 is a natural product (of course) isolated from sponge Pseudoceratina sp.

Well, that’s not very complex target and the retrosynthesis and also total synthesis should be quite easy. My idea to disconnect this stuff is something like that:

Everything should by clear. All starting materials could be prepared from phenole or from commercialy avaiable anisaldehyde.

People who did the synthesis used very simmilar approach, so I didn’t draw retrosynthesis again. One important difference is use of Perkin condensation instead of something I labeled ‘Wittig-like’ reaction Let’s see how they completed whole total synthesis.

As you can see they started with p-anisaldehyde and they brominated it in first step. Ok, reaction is slow (there’s no Lewis acid), but more interesting is that aldehyde functional group isn’t oxidizied under such conditions. Coversion 3->4 is of course mentioned Perkin condensation (well, authors wrote that is Doebner-Knovenagel condensation… ok, probably it is ). Last step in this scheme is conversion of carboxylic acid to corresponding acyl chloride.

Now, second building block is prepared. They use p-hydroxy benzyl alcohol (can be obtained from p-anisaldehyde by its reduction) as second precursor. Next substitution oh hydroxyl group is performed to give 7. It’s a question why they didn’t convert -OH to better leaving group such as tosylate and so forth. 7 is reduced on palladium catalyst to amine hydrochloride 8, which in turn is brominated to 9 (again without any Lewis acid).

In next step two building blocks 5 and 9 are coupled and amide 10 is formed. OH group from aromatic ring of 9 isn’t acylated.

10 is a core of aplysamine 6. In next two steps primary amine moiety is attached to this structural core. This is achived by use of Williamson reaction (ether formation) with 1,3-dibromopropane to form derivative 11. 11 is converted to aplysamine 6 by well-known Sn2 reaction.

For more pieces of information please see:

N. Ullah, K. M. Arafeh, Tetrahedron Lett., 2009, 158-160.

Do you know snowdrops? It’s well-known that bulbs of these flowers (latin name is Galanthus nivalis) contain many alkaloids and galantamine (or galanthamine) is one of them:

This is an important natural product because of its biological properties and phamacological applications – it’s used in treatment of mild Alzheimer’s disease. So there are many approaches to total synthesis of galantamine and here I’ll try to show most recent of them (I think so).

Authors of the paper on which I base developed new interesting reaction: DMCRC – what means Double Michael-Claisen Reaction Cascade. The reaction allows to synthesise quickly highly substituted cyclohexenones which can be used in total syntheses of many ’sterically congested’ natural products and galantamine is only one of several examples mentioned in paper (the others are aspidospermidone, lycoramine and  mesembrine).

Let’s look at retrosynthetic analysis:

As you can see, that organic synthesis of galantamine starts with arylated acetone. Now, let’s see how it was acomplished:

Mentioned before acetone 2 was undergone DMCRC (yeah, exercise that name one more time – Double Michael-Claisen Reaction Cascade) reaction with tert-butyl ester of acrylic acid. The mechanism of this conversion isn’t so obvious and you can find full explanation (with some calculations of transition states) in paper. The most important thing is that termodynamic enolate of 2 reacts faster with acrylic ester than kinetic enolate of 2. This is the secret of this reaction

Let’s get back to synthetic route. Formed 1,3-dienone 3 is converted in next step to enol ether 4 which is next reduced to enone 5. Enone 5 is then protected (self-protected) by primary alcohol moiety in Michael-type reaction and this allows to selective removal benzyl group to give 7 without any saturation on carbon-carbon double bond.

Then released phenolic -OH group participates in five-membered fused ring and 8 is formed. Next, two oxidations were performed to oxidise primary -OH group to carboxylic acid. Transformation 9 to 10 is a Curtius rearrangement and DPPA (DiPhenyl PhosporoAzidate) is a donor of azides here. Now, Pictet-Spengler cyclization occurs to give 11, and mechanism of this reaction is drawn below:

Synthesis of galantamine is completed in such way:

There is nice method of conversion cyclohexanone 11 to cyclohexenone 13 in palladium-catalyzed process. Mechanism is:

β-elimination of organopalladium compound can only occur at one side of carbon-oxygen double bond.

In last step 13 is reduced by L-Selectride (stereoselective reduction of carbonyl group) and LiAlH₄ (reduction of ester moiety) and galantamine 1 is formed.

For more pieces of information of course see:

T. Ishikawa, S. Saito et al., J. Org. Chem., 2008, 7498.

You can read here today about total synthesis of (+)-Isofregenedol and, in my opinion, this is the very total synthesis which looks like a magic. In one aspect, at least.

Let’s see our current target molecule:

It looks interestingly. Only one chiral carbon atom, but five-substituted aromatic ring – in other hand. This stuff is a diterpene which was isolated from chilean flower Haplopappus parvifolius by Niemeyer in 1991. (+)-Isofregenedol exhibits important biological activites. It’s – for example – methionine aminopeptidase-2 reversible inhibitor, agonist of glucocorticoid receptor and agonist of serotonin 5-HT₇ receptor. Sounds seriously.

Before you’ll see how this molecule were prepared, I’ll show you my approach to its organic synthesis. First – retrosynthesis:

As you can see, my synthesis would be start with bromoxylene. However, I know that there could be some problems with regioselectivity during Friedel-Crafts acylation step. But see:

I think it can work.

Ok, let’s leave this and deal with real organic synthesis of (+)-Isofregenedol done by authors of paper. Their total synthesis has a few more steps than my, but theirs really was done So let’s look at retrosynthetic plan:

Lots of disconnections. And one of more important disconnection here is gold(I)-catalyzed benzannulation. This is that magic about which I write. Some details are shown below:

Initially authors wanted to complete synthesis in very effective way. They wanted to perform benzannulation of highly substituted cyclohexane ring (obtained from geraniol as chiral building block), but unfortuantely such reaction gave no product. So plans had to change. Modified approach is shown below in all its details:

Total synthesis of (+)-Isofregenedol starts with cyclohexene epoxide 2. 2 is undergone epoxide ring opening reaction with vinyl Grignard reagent and followed by Swern oxidation to give 3. Next, addition of alkynyl Grignard reagent to ketone 3 occurs and tertiary propargyl alcohol 4 is formed. Everything is ready to carry out gold(I)-catalyzed benzannulation reaction:

Yield of the reaction is 69%, but in this step aromatic ring with four substituent is created. But how it really works? One of possible mechanisms is drawn below:

When I look at this mechanism then everything becomes so clear

Ok, so we have already tetrahydronaphtalene core 5 of (+)-Isofregenedol, now we have to link to it all lacking substituents.

Conversion of 5 to 7 maybe very elegant. Probably regioselctivity of benzylic oxidation isn’ very good. But addition of dimethylzinc in the presence of TiCl₄ as a Lewis acid looks nice. And in such way we have dimethyl derivative 7 as major product (under standard condition we rather would expect tertiary alcohol). Now, 7 is converted with high yield (excellent regioselectivity) to 8 and then Stille coupling is performed (some microwave chemistry aspect) and vinylbenzene 9 is formed. In next step double bond of 9 reacts with 9-BBN in hydroboration reaction to give intermediate 10.

This intermediate is combined with iododerivative 11 under Suzuki coupling conditions. Compound 12 is formed and it’s reduced with DIBAL-H to give allylic alcohol 13. In next step enantioselective Sharpless epoxidation is undergone (with excellent yield and high enantiomeric excess) and epoxide 14 is formed. In next two steps which include formation of iododerivative and epoxide ring opening (+)-Isofregenedol is obtained. And total synthesis (consisting of 13 steps and where no protecting groups was needed) is completed.

Fore more – as always see:

M. Riou, L. Barriault, J. Org. Chem., 2008, 73, 7436.