A Concise Synthesis of (−)-Lasonolide A

A Concise Synthesis of (−)-Lasonolide A

Barry M. Trost, Craig E. Stivala, Kami L. Hull, Audris Huang, and Daniel R. Fandrick

DOI: http//dx.doi.org/10.1021/ja411270d

Not so many total syntheses have been published these days but this one caught my attention (some might say for obvious reasons…). Though lasanolide A has been made a couple of times but never in such a neat fashion utilizing some pretty efficient metal catalyzed processes. Trost’s retrosynthesis is shown below. The two major fragments were assembled by an intermolecular ruthenium mediated enyne coupling and a Yamaguchi macrocyclization. The western fragment in turn derives from an alkyne precursor to which the side chain is attached by consecutive HWE and Wittig olefinations. The stereochemistry is set by a highly efficient ProPhenol aldol reaction. The eastern fragment also makes use of a HWE olefination and a Hiyama coupling, respectively. A ruthenium catalyzed hydrosilylation and cross metathesis then give retrosynthetically the diol shown whose stereochemical information derives from an enzymatic reduction and an old-school CBS reduction.

 Scheme 1

  scheme_1_19012014

Starting in the forward sense the side chain of the western fragment was prepared in just four steps. Neopentyl cuprate addition to propargyl alcohol was followed by acetonide cleavage / esterification. TBS protection and formation of the Wittig salt furnished fragment 3 in a pretty efficient manner.

 Scheme 2

 scheme_2_23012014

The other part of the western fragment originates from a ProPhenol mediated aldol reaction to give 6 in nearly enatiomerically pure form. DIBAL reduction and subsequent TBDPS protection was followed by stereoselective acetal formation and Ley oxidation yielding 8. Only one diastereomer is obtained in the acetalization step under thermodynamic conditions. Deprotection of the alcohol directly provided the hemiacetal 9 as an inconsequential mixture of diastereomers. HWE olefination and DIBAL reduction produced aldehyde 10 which could be coupled with 3 in the presence of KHMDS. After a survey of methods the acetal cleavage was preferably accomplished with LiBF4 completing the synthesis of the western fragment.

 Scheme 3
 scheme_3_23012014

Ok, now let us have look at the synthesis of the eastern fragment. It kicks off with an old school Blaise reaction (Reformatsky with a nitrile) to give dicarbonyl 12 which eventually was reduced to hydroxyester 13 by an enzyme called CDX-024. TIPS-protection and two-step conversion of the ester to ketone 15 was followed by a these-days-rather-rare CBS reduction to propargyl alcohol 16. Ensuing trans-selective hydrosilylation developed in the Trost labs and cross-metathesis with crotonaldehyde furnished directly pyran 18.

 Scheme 4

scheme_4_23012014

The remaining acid side chain was introduced by HWE olefination using lithium hydroxide in the presence of mole sieves. Allylation utilizing a Hiyama-coupling of the vinyl silane, saponification and TBS protection then furnished eastern fragment 22.

 Scheme 5

 scheme_5_23012014

Completion of the synthesis was brought about by a powerful intermolecular ruthenium mediated enyne coupling giving mainly the linear isomer 23 in a 3 : 1 ratio. This result is rather unprecedented because this reaction usually favors the formation of branched products. As acetone proved to be the most effective solvent in this reaction under the conditions employed the formation of an acetonide was observed which was cleaved off with CSA in an ensuing step. After selective TBS-protection of the three least hindered alcohols the seco acid was closed under Yamaguchi macrolactionization conditions to give after global deprotection lasonolide A in an overall yield of 1.6 % (16 steps LLS with respect to allyl cyanide).

A really impressive synthesis relying mainly on powerful transition metal catalyzed transformation. I hope you also liked the quiz… If you have any suggestions please let me know.

 Scheme 6

 scheme_6_23012014

Advertisements

Total Synthesis of the Galbulimima Alkaloid (-)-GB17

Total Synthesis of the Galbulimima Alkaloid ()-GB17

 Reed T. Larson, Michael D. Clift, and Regan J. Thomson

 DOI: http://dx.doi.org/10.1002/anie.201108227

GB 17 belongs to the family of Galbulimima alkaloids which can be found in the bark of a rainforest tree with himbacine as a promising lead structure for muscarinic receptor antagonists. Himbacine-like compounds were tested for the treatment of Alzheimer’s disease as thrombin agonists.

Other family members including himandrine, GB13, himgaline, and GB16 have been synthesized. To date no synthesis of GB17 is known so the Thomson group accepted this last challenge. The retrosynthetic analysis is shown below. Nothing real spectacular but a nice access to the tetracyclic carbon skeleton is presented.

 Scheme 1

 

The first building block is readily available by a methodology developed by Lhommet et al. Reaction of ester 1 with (S)-phenylglycinol yielded oxazolidine 2 which was hydrogenated to give piperidine 3. The yields are not reported but considering the original publication about 40 % yield can be achieved.

The linchpin was synthesized starting from monoprotected diol 4 which was converted to iodide 5 and dithiane 6 which alkylation with 7 to give acetal 8. 5 and 8 were coupled and the aldehyde and alcohol were freed with dilute HCl in acetone. [3]

 Scheme 2

Treatment of ester 3 with lithiated phosphonate and Boc protection of the naked amine gave ketone 10 [4]. HWE reaction with linchpin 9 under Masamune-Roush conditions and subsequent DMP oxidation furnished aldehyde 11. After some model studies the group found that a TMS-prolinol catalyst gave highest yields and enantiomeric excess on a multigram scale. In a one-pot procedure the aldehyde was converted to unsaturated ester 12. Base induced cyclization, amine deprotection, and lactamization yielded tetracycle 13 in moderate yield. Nevertheless it was found that the wrong isomer had been formed together with complete inversion of the stereocenter next to the amine.

Scheme 3

 

Obviously the (E)-configured ester gave the wrong stereochemistry in the Michael addition step, so the group proceeded from 11a through a Still-Gennari modified HWE to give again under Masamune-Roush conditions ester 14. Boc deprotection and this time sodium methanolate induced cyclization did the job. Under these conditions the lactamization occurred to give 15. The keto group in 15 was removed under standard conditions by formation of the vinyl triflate which was reductively removed in the presence of Pd and formic acid as the hydrogen source. Stereoselective alkylation of the lactam was followed by dithiane removal, reduction, and oxidative cleavage of the exo-methylene group to give GB17.

Scheme 4

 

To explain the stereochemistry in the organocatalytic step I would propose the following transition state. Enamine formation of the prolinol ether should lead to the transition state with the least steric interactions. [5] McMillan’s catalyst or proline gave much lower ee values.

 Scheme 5

 

The outcome of the cyclization step can be explained considering the transition states shown below. In structure 12 steric interactions between the large Boc group and the ester force the double bond into an axial position. Alternatively without the Boc group and with a (Z)-double bond the ester group is equtorial so steric interactions can be minimized in the conformer shown.

 Scheme 6

And as usual THX to Bobby for proofreading.

[1] DOI: http://dx.doi.org/10.1016/j.tet.2005.05.079

The specificity of the reduction step can be explained by looking at the particular bonds which are reduced. 1) The enamine bond is reduced stereoselectively by facial differentiation from the Re face. 2) The aminal opens up to an imine which is again reduced to the amine. 3) The auxiliary is cleaved off.

[2] DOI: http://dx.doi.org/10.1021/ja055740s

[3] Have a look at BRSM’s blog for a sweet discusison on linchpins…

[4] Interestingly the group protected the amine after the BuLi chemistry which results in the usage of > 2 eq of lithiated phosphonate. Maybe earlier Boc-protection gave racemisation through DoM-chemistry with some help from the Boc group. Racemization was later found to occur in the presence of tBuOK.

[5] The group stated that the dithiane protecting group was essential for the reactivity of the substrate. Without this group almost no transformation was observed. Considering the great Thorpe-Ingold effect of this protecting group it might be an explanation.

A Novel Approach to Indoloditerpenes by Nazarov Photocyclization: Synthesis and Biological Investigations of Terpendole E Analogues

A Novel Approach to Indoloditerpenes by Nazarov Photocyclization: Synthesis and Biological Investigations of Terpendole E Analogues

Fa´tima Churruca, Manolis Fousteris, Yuichi Ishikawa, Margarete von Wantoch Rekowski, Candide Hounsou, Thomas Surrey, and Athanassios Giannis

DOI: http://dx.doi.org/10.1021/ol100579w

As the title suggests it’s time for some sunlight chemistry… Ok only one step but the rest of the synthesis is also worth reading. The Terpendoles are a family of indoleterpenes which show weak activity as acyl-CoA:cholesterol acyltransferase inhibitors. Recently it was discovered that the terpendoles inhibit the kinesin spindle protein (KSP).

In this paper the synthesis of one member of this class is described. The retro is rather short as the paper is, too. We’re starting with some FGI and cut the molecule into two halves by using the above mentioned Nazarov cyclization strategy. As can easily be seen, the molecule should be accessible directly from the known Wieland-Miescher-ketone.

Retro:

So here we go:

The scheme starts with a selective protection of the unconjugated carbonyl. Phenylthiomethylation (search for Kirk-Petrow-reaction for further information) which was followed by a SET reduction under Birch conditions and subsequent trapping of the anion with allylbromide then yields the allylated/methylated ketone. LAH reduction of the remaining ketone, boronation of the terminal olefin and oxidation results in lactone formation. Oxidation of the ketone lactone to an α-β-unsatured one was achieved under more or less unconventional conditions. Epoxidation with H2O2, epoxide opening with phenylselenide and protection of the resulting alcohol as the MOM ether closes the first scheme.

Scheme 1

Because the phenylthiomethylation looks a bit odd, here is the mechanism:

Mechanism 1

The first few steps should be clear. The Birch reduction step might involve an intermediate radical anion which is trapped by allylbromide and reprotonated under thermodynamic control.

Furthermore the dehydration dehydrogenation step with this to me unknown reagent:

Mechanism 2

This is only a proposal of what I think the mechanism might be… I’m open for better ideas or corrections.

With the blue intermediate in hand we can move on. Selective reduction of the lactone was achieved with DIBAL-H and the aldehyde olefinated. Epoxidation of the alkene with mCPBA was followed by Sc(III) mediated pyran formation, oxidation and epimerization of the isomeric ethers to give one single pyran ring. Grignard reaction with methylmagnesium chloride, PG interconversion and acetal cleavage sets the stage for the final few steps.

Scheme 2

The first step involves an aldol condensation/hydrogenation to link both halves of the molecule together. Methylation of the ketone, benzylic oxidation with DDQ, dehydration of the tertiary alcohol with Burgess reagent to the exocyclic alkene and isomerization of the latter one to the endocyclic alkene prepares the key intermediate for the Nazarov cyclization. This [2+2] cyclization was mediated by UV light and closes the ring in a disrotatory manner.

Scheme 3

Protecting group removal and complete reduction of the ketone then yields Terpendole E.

Scheme 4

Overall a nice synthesis but I would have preferred a bit more details in the paper. The authors only give the used reagents without any more information like conditions or eq’s. Nevertheless nice chemistry but there’s a little mistake in the published paper. Maybe you find it too…

Any comments?

Asymmetric Synthesis of (+)-Polyanthellin A

Asymmetric Synthesis of (+)-Polyanthellin A

Matthew J. Campbell and Jeffrey S. Johnson

DOI: http://dx.doi.org/10.1021/ja904136q

Hell yeah, back from the grave…

It took some time but now it’s finished: this time featuring a very cool asymmetric synthesis of the natural product Polyanthellin A. I do remember an extremely effective synthesis from Overman in this class of natural products a few years ago employing his oxy-Cope/Mannich-tandem reaction. In this paper the attention is less methodical nature and more focussed on the target itself.

Whatsoever let’s get started:

If you’re interested in the biology of this little metabolite take a look in the original report in Natural Products (ref. 7)…

First the retro:

polyanthellin A_160909

The key step in this synthesis features an asymmetric formal [3+2] cycloaddition starting from the 2 fragments which I will call from now on the red and the blue one.

This paper is full of interesting chemistry, too much for this little review, so I will focus only on the key aspects. If you’re interested we can discuss the reactions I did not picture in full detail later in the comments.

The blue fragment was synthesised starting from commercially available methallyl alcohol:

Scheme 1

scheme_1_160909

Sharpless asymmetric epoxidation creates the only stereocenters in this fragment, followed by an epoxide opening from a copper catalysed allylation. Chemoselective tosylation, Kolbe-Schmitt-nitrile synthesis, TMS protection and DIBAL-H reduction furnished the blue fragment in an overall acceptable yield (the solvent is choice is odd, I would have taken THF)…

Synthesis of the red one needed some more attention but in the end makes use of a bunch of quite efficient stereoselective methods. And here it comes:

Scheme 2

scheme_2_160909

First a more or less standard Michael addition catalysed by Prolinol-derivative (1) and catechol ester (2) to give the 1,5-diketone in high yield. A Wittig reaction with titanated allyldiphosphine yields the required Z-allyl side chain which was followed by methylcarboxylation using Mander’s reagent furnishing the functionalized malonester. Diazotransfer and subsequent carbene inserton catalysed by (3) into the nearer double bond gives in the end the red fragment ready for the formal [3+2] cycloaddition.

Now the key step: After an extensive screening of catalysts the authors found this bulky aluminium based Lewis-acid catalyst giving the best results. The complete scheme looks like this:

Scheme 3

scheme_3_160909

However they needed 3eq of the blue fragment but this gives them the core structure in a very good yield and stereoselectivity. A metathesis employing Grubbs II closes the ring ensued by Krapcho decarboxylation, a sequence of hydroboration/TPAP oxidation and another Wittig reaction completes this scheme.

Because there are no mechanisms in this paper I created this one for the [3+2] cycloaddition:

mechanism_160909

First a Lewis-acid catalysed cyclopropane opening gives the stabilised allyl cation and the enol ester which in turn attacks the carbonyl in an aldol fashion followed by ring closure from the enolate oxygen. Or a more concerted cyclization?

Ok, nevertheless only a few more steps to go:

Scheme 4

scheme_4_160909

Simple iodo etherification, oxymercuration and global reduction yields the naked Polyanthellin which was acetylated to give the desired product.

To my surprise the JACS paper is only 2 pages long… I would have expected the paper to be at least 4 to 6 pages long to show how they employed the specific methods cause some of them a really rare. That’s why I prefer the Angewandte papers: they feature almost the complete synthesis in detail, which is not always useful but gives you a better insight into the planning and realization of such a complex synthesis.

Ok, that’s it from my site, any comments?

___________________________________________________________________________________________

To the selectivity in the iodoetherfication:

I made this nice little 3D model with ChemDraw which might explain the differentiation. The cyclohexane-methylen group is blocked from equatorial attack by the hydroxy functionality cause the cyclononane ring forms a twist boat whereas axial attack is blocked as usual by  1,3 axial strain.  Or the access to the other methylen group is much easier as we’re dealing with a more convex shape of the molecul in this area. Here is what I mean: the green ring marking the blocked methylen-group, the red ring marking the attacked one:

iodoetherfication

But yeah, cool selectivity!! 🙂