Extreme Diels-Alder Reactions Your Chemistry Teacher Warned You About

[00:00:00]Even if you re not a spongiologist and yes that s in fact a real job you should appreciate the ugly inconspicuous sponges during your next diving trip.

[00:00:09]They might contain valuable natural products like lissodendoric acid A, dubbed by some a potential anti-Parkinsons medicine.

[00:00:17]Why would a random Spongebob produce this complex thing?

[00:00:20]We ll get to that in a moment.

[00:00:22]Maybe you re thinking hey man, the title says Diels-Alder reaction, so where is the keying element in the molecule?

[00:00:28]Well, only the chemistry grand-wizards will know that there is indeed an uber exotic Diels-Alder hidden in this structure.

[00:00:36]In today s video, we will explore two truly extreme total syntheses, one of them featuring an insane quadruple Diels-Alder cascade.

[00:00:45]This system was so hungry for Diels-Alders that the authors literally had to stop it from going quintuple.

[00:00:52]We will cover basic and pretty advanced organic chemistry, offering a nice learning opportunity for everyone so no one should be diene today.

[00:01:01]So to those of you unfamiliar with this type of seemingly random and pointless natural product research, it basically all starts with digging around in the sea for your specimen these sponges seem to like it particularly cold.

[00:01:15]After some marine dredging, a.k.a.

[00:01:16]playing Bob the Builder underwater, you grab your sponge and bring it to your trusted spongiologist to validate its authenticity.

[00:01:24]After this, you finely chop it up like Gordon Ramsay ordered you to and then do a bunch of chromatographic steps.

[00:01:35]To the trained eye, this seems like an easy-mode isolation compared to something ridiculous like ciguatoxin.

[00:01:41]As I ve shared in a by now almost ancient video, those chemists spent some fun weeks with 4 tons of moray eels and a bunch of blenders.

[00:01:47]To keep track of the trace toxin during the isolation steps, they simply checked which fraction killed their test mice.

[00:01:54]All of that work was reworded with half a milligram of natural product.

[00:01:58]In comparison, cutting up 200g of sponge and running some columns and HPLC seems like heaven.

[00:02:05]So after analyzing every fraction for unknown compounds, you eventually find a new gem like lissodendoric acid A. To characterize it, you try to draw as many NMR correlation arrows as humanly possible, run some mass spec, and so forth.

[00:02:20]This research also includes defining and more rarely testing hypotheses on biosynthetic origins of the product sometimes, these include quite exotic reactions.

[00:02:31]The scheme here we ll not go through it in detail was actually drawn based on what s thought about a related natural product, manzamine A.

[00:02:39]This thing was first isolated in 1986 and it has since attracted attention of chemists, but it s biosynthetic origin is still a mystery four decades later.

[00:02:49]The original hypothesis by Baldwin had some experimental evidence behind it, but it envisioned a multi-component condensation of different educts.

[00:02:58]These building blocks are a bit dubious to say the least.

[00:03:01]Acrolein is usually generated as a by-product of catabolic processes and is rapidly metabolized to less toxic species.

[00:03:09]The supposed dialdehyde is extremely reactive and doesn t have a biological function.

[00:03:14]Ammonia is obviously commonly encountered, however not in its free form.

[00:03:19]So, still in 2023, there are new suggestions to the first few steps of this cascade, including the postulation of free fatty acids as initial building blocks.

[00:03:30]This might be sensible because many different cousins of manzamine feature varying chain lengths and most commonly (Z)-alkenes, the most predominant form of unsaturated fatty acids.

[00:03:40]Just like most toxins, it s actually not the sponge who produces manzamine so these biosynthetic pathways occur in sponge-associated bacteria.

[00:03:50]For instance, researchers found a strain M42, capable of producing manzamine A however, only in small quantities at a specific time early in the growth cycle.

[00:04:01]Behind this, there is a weird symbiotic relationship of sponges tapping into a chemical defense mechanism and in return offering a controlled habitat for bacteria.

[00:04:11]The same phenomenon is behind more famous toxins like tetrodotoxin which is synthesized by bacteria and only in a second step sequestered or accumulated by puffer fish and friends.

[00:04:22]Ok, that s enough biology for today.

[00:04:25]The last step of every isolation is testing the natural product on a bazillion different potential properties.

[00:04:30]In this regard, my favourite lackluster activity is the ability to kill amaranth and lettuce.

[00:04:36]Fortunately, lissodendoric acid A had more impressive effects in a validated in vitro model of Parkinson s disease.

[00:04:43]Here, the molecule protected nerve cells from oxidative damage at very low concentrations.

[00:04:49]It s not a breakthrough by any means but the hope is that this molecule could inform the understanding and design of future Parkinson s medicines.

[00:04:58]The value of the synthesis we will discuss is not about facile access to these biochemical properties but rather due to the application of a new methodology.

[00:05:07]Retrosynthetically, there are several ways to deconstruct the target.

[00:05:10]If you have some good chemistry knowledge already, you can think about this problem for some minutes yourself.

[00:05:16]Drawing things on paper is much easier than finding what works in real life, so some of you will draw syntheses that are even simpler than what we will discuss.

[00:05:24]Because this approach centred on the Diels-Alder reaction of a cyclic allene?

[00:05:31]Most chemists know allenes from the time they got bombarded with name reactions to learn by heart.

[00:05:36]These species are useful for niche applications I m thinking of some freaky gold-catalyzed rearrangement cascade here and indeed, have also been employed in cycloadditions, typically 2+2 or Pauson-Khand reactions.

[00:05:50]Cyclic allenes have also been known for decades.

[00:05:53]However, most research is limited to obscure compounds like this one.

[00:05:57]Ironically, this 10-membered ring allene can be made with the Doering-LaFlamme synthesis so do not forget to learn your name reactions and mechanisms!

[00:06:06]If you were wondering, this compound in particular was used to study mercury-catalyzed rearrangements.

[00:06:12]The research at hand upped the ante by leveraging a strained six-membered cyclic allene.

[00:06:19]The groundwork for this had been laid by the research group a few years ago already, as they identified suitable chiral silyltriflate and silylbromide precursors which upon treatment with cesium fluoride in acetonitrile generated highly stereospecific allenes.

[00:06:36]The strain energy is estimated to be in the ballpark of 30 kcal/mol so even though they look ultra-weird, they are less strained than benzyne, for instance.

[00:06:46]Still, they are highly reactive and axially chiral, so they were able to trap them in diastereoselective Diels-Alder reactions.

[00:06:52]The interesting thing is: we have two di-enophilic groups!

[00:06:57]I never I thought I would use this as an adjective.

[00:07:00]Using the typical electron-rich diene furan, this addition is regioselective for the less substituted double bond, with some erosion of enantiomeric eccess.

[00:07:10]Regarding diastereoselecitvity, note that the addition occurs from the top face, even though you might think that the methyl group is directing the selectivity.

[00:07:18]Also, the reaction is selective for the endo product as C-H and C-O furan bonds are pointing to the same side.

[00:07:26]We will look at the transition state shortly to better explain this.

[00:07:30]This reaction is cool, but for the product to be synthetically useful for lissodendoric acid, we require the opposite regioselectivity.

[00:07:38]A key insight here was that the use of an electron-poor pyrone diene gave the desired connectivity with even better ee.

[00:07:47]This selectivity is in line with what you would expect from an inverse-electron demand Diels-Alder, with the more substituted dienophile being more reactive.

[00:07:56]You can check for yourself that it s still endo selective, and that the diastereoselectivity is directed by the methyl group this time.

[00:08:03]The pyrone will be crucial for the total synthesis as we will see.

[00:08:08]If you wonder why there is an erosion of ee this comes from racemization of the fleeting allene.

[00:08:14]Based on the substituent, the energetic barrier is higher or lower, thus influencing the degree of ee transfer.

[00:08:22]You might want to say bro, just use a normal olefin instead of the allene?

[00:08:25]Why are you overcomplicating it?

[00:08:26]That s exactly what the authors tried initially with exhaustive screening, so opting for this very exotic method instead was justified.

[00:08:35]Let s check out how the total synthesis played out.

[00:08:39]The pyrone diene was easily prepared from this commercial carboxylic acid.

[00:08:44]A double tosylation and esterification gave the activated tosylate which was Negishi-coupled to a long alkyl rest.

[00:08:52]You will see that it bears a terminal alkene which will ultimately be handy for the synthesis of the macrocycle.

[00:08:59]Note it differs from the model substrate we saw because that one had a benzylated hydroxy instead of the protected ester.

[00:09:06]The allene precursor was a bit more challenging to synthesize.

[00:09:11]Analogous to the diene, a long alkyl rest was first introduced by treatment of this starting material bearing a sp2-triflate with a zirconium reagent and copper with the bromide group chilling due to its lower reactivity.

[00:09:25]This alkyl rest will direct our allene cycloaddition later on.

[00:09:29]To selectively introduce the silyl group required for the allene generation, 3 steps were needed.

[00:09:35]First, chirality was introduced through a CBS reduction of the carbonyl.

[00:09:40]After activation with ethylchloroformate, the substitution with asilyl cuprate nucleophile gave the allene precursor, proceeding with perfect SN2 stereoinversion.

[00:09:50]For the Diels-Alder reaction, they found that -20 C in acetonitrile was the optimal temperature to maximize transfer of chirality.

[00:09:59]This required the use of tetrabutylammonium bromide as a phase transfer catalyst to increase solubility of the caesium fluoride required to trigger the allene formation.

[00:10:10]We ve already covered that there are many different potential products impressively, the desired isomer was the sole observable product of this reaction.

[00:10:18]We ve already indicated that the reaction follows endo selectivity here, it might help to think of the non-reactive olefin at the allene as other electron-withdrawing groups.

[00:10:30]There is some overlap between the orbitals of the diene with that pi-system, which again explains why the R1 group and the ester are syn in the product.

[00:10:40]The regioselectivity on the other hand can be explained by looking at the resonance structures of the educts.

[00:10:45]To complete the total synthesis, a concise late-stage sequence was established.

[00:10:50]First, they had to get rid of the non-reactive allene olefin.

[00:10:54]Because many conditions for this reduction did not work, they had to take a detour by first oxidizing the alpha-amino carbon to the unsaturated lactam, and then throwing a copper hydride on it.

[00:11:05]Here, we should highlight the value of the pyrone Diels-Alder fragment.

[00:11:08]Through the resulting CO2 bridge, the 1,3-diene motif of the natural product is masked it would be impossible to reduce the desired double bond selectively if it would be conjugated with the diene as well.

[00:11:22]After the bridge served its purpose, the team removed it via thermal decarboxylation, used a copper-mediated Boc deprotection to release the free nitrogen, and appended another rest bearing a vinyl group.

[00:11:34]You probably now realize that an olefin metathesis reaction can link things up with the vinyl group which was chilling on the sidelines all along.

[00:11:42]Now, using a Rhodium-catalyzed reduction protocol, they were able to selectively target the imide while keeping the diene and ester groups intact.

[00:11:51]Last, Boc deprotection with acid gave lissodendoric acid.

[00:11:54]With just shy of 1% overall yield, this will not pave the way for major medicinal investigations but the methodology absolutely qualifies as an extreme Diels-Alder in my book.

[00:12:06]Did you catch a new insight or learning yet like Spongebob during jelly fishing?

[00:12:10]If so, please show your support by liking, commenting and subscribing.

[00:12:14]A massive thanks goes to my channel members on Patreon and Youtube for supporting my content.

[00:12:20]I m really grateful for your encouragement, and any comments or feedback you might have regarding future videos.

[00:12:27]Equally extreme as our first topic are the reactions part of the total synthesis of pedrolide.

[00:12:33]This natural product has no nice sponge story because it is a bit less exciting, originating from Euphorbia plants they are the sources of many many terpenes with interesting bioactivities and structures.

[00:12:45]Following its isolation in 2021, this natural product was shown to inhibit p-glycoprotein membrane transporters, which are connected to drug resistance of cancer cells maybe cool, but nothing groundbreaking, and definitely not as flashy as anti-Parkinson activity.

[00:13:03]What this molecule lacks in function, it makes up in complexity.

[00:13:08]The extreme Diels-Alder reaction in question is critical to assemble this impressive structure.

[00:13:14]But similar to Lissodendoric Acid A, we have to ask ourselves where could we envision a Diels-Alder retron?

[00:13:20]Obviously, we should start to look at the six-membered rings in the structure, and understand if we can reverse engineer a cyclohexene motif somewhere.

[00:13:29]The left part is particularly inviting because it s highly interconnected.

[00:13:33]Let s assume that the alpha-methyl ketone could be synthesized from the corresponding double bond and voila, we have a cyclohexene.

[00:13:41]If we break the ring apart, we realize that this would be, in contrast to lissodendoric acid, a normal-electron demand Diels-Alder reaction based on cyclopentadiene and an alpha-beta unsaturated lactone as an electron-poor dienophile.

[00:13:56]This conformation might look weird, but redrawing helps to clarify.

[00:14:00]How can we embed the two Diels-Alder educt systems into the molecule?

[00:14:04]Let s take the dienophile.

[00:14:06]This alpha-beta unsaturated lactone might originate from an elimination reaction this approach is smart because it simplifies the linear chain s introduction earlier in the synthesis, as we will see shortly.

[00:14:17]The simplest disconnection of a lactone is simply an intramolecular lactonization, requiring a free hydroxyl and the carboxylic acid.

[00:14:26]What about the diene?

[00:14:27]It s greyed out here as it was not used in the actual synthesis like this.

[00:14:31]That s because we need so tricks up our sleeve to make chemistry with this group.

[00:14:36]On one hand, cyclopentadiene is really reactive.

[00:14:39]Why?

[00:14:40]First, due to its cyclic nature, it s permanently locked in the s-cis conformation required for the concerted Diels-Alder mechanism.

[00:14:48]Second, a lower distortion energy required to reach the transition state geometry makes it also more reactive than other rings, such as cyclohexene.

[00:14:57]However, the high reactivity also comes with unproductive isomerization through sigmatropic hydrogen shifts, dimerizations and other reactions.

[00:15:08]A synthetic study which exemplifies this was focused on intramolecular Diels-Alder reactions.

[00:15:11]They found that their original cyclopentadiene isomerized in situ, leading to a quite different product than what they envisioned.

[00:15:20]Even though we might expect the more substituted isomer to be most stable, we can always have Curtin-Hammet situations which mess up our plan.

[00:15:28]This high reactivity needs to be tamed so cyclopentadienes need to be masked throughout the synthesis, and unveiled immediately prior to their use.

[00:15:38]Although other masking groups were considered, norbornadiene was used as a surrogate in this synthesis due to its stability and mild release conditions you will understand this once we check out the cascade in detail.

[00:15:51]Another benefit of norbornadiene is that it can be introduced with a nucleophilic addition to the carbonyl.

[00:15:57]To break down the intermediate further, we can assume that the carboxylic acid could come from oxidation of the primary alcohol, and that the whole side chain can be appended to the six-membered ring through enolate chemistry.

[00:16:10]This educt does have a dimethyl cyclopropane group which is present in many commercially available terpenes so you might think that the synthesis taps into nature s chiral pool.

[00:16:21]But due to the specific functionalization pattern, the chemists instead assembled the thing bottom up.

[00:16:27]It all starts with the oxidative de-aromatization of 4-methoxyphenol.

[00:16:31]The chiral methyl group was then introduced through a conjugate addition with enantioselectivity directed by the well-known phosphoramidite Feringa ligand.

[00:16:40]Then, a Rubottom oxidation via the silyl enol ether introduced an alpha-hydroxyl group.

[00:16:46]So we have started to decorate our central six-membered core with the methyl and hydroxyl group in place but what about the dimethylcyclopropane?

[00:16:52]First, the necessary carbons were introduced through Grignard addition to the ketone.

[00:16:58]There was a good diastereoselectivity for the desired anti-diol, likely through the adjacent hydroxy group which directs the addition from the same side.

[00:17:07]Also, note that because a normal Grignard reagent was used, 1,2 addition was preferred as opposed to conjugate 1,4 addition.

[00:17:15]Next, the dimethyl acetal was hydrolyzed as we will need the ketone shortly, and the diol was protected with TES groups.

[00:17:22]To forge the cyclopropane, the newly introduced isopropene unit was reduced and coupled through a 1,4 addition.

[00:17:29]It s quite impressive that even though the olefin is not activated, the reaction proceeds in very high yield.

[00:17:36]Alright, so next up we have to work through what we previously simplified as enolate chemistry.

[00:17:42]As the authors described, initial attempts at side-chain installation with LDA or LiHMDS resulted in low yields.

[00:17:50]Interestingly, doing it in an indirect manner worked much better first converting the ketone to the TMS enol ether, then generating the lithium enolate by exchange with methyl lithium, and only then running an aldol reaction with this aldehyde bearing the side chain.

[00:18:06]Here we can see the random nature of experimental chemistry because the free beta-hydroxy group was unstable, an immediate silylation was used to protect the product.

[00:18:17]This worked well at small scales but with increasing quantities, the protection was more and more sluggish.

[00:18:22]For whatever reason, running the reaction at 2 mmol scale led to no silylation at all.

[00:18:28]Things like this just make your blood boil.

[00:18:31]Next up, they performed the nucleophilic addition of the norbornadiene organolithium.

[00:18:36]In the literature, this group had previously only been added to more reactive aldehydes so it was likely quite a relief to see that the more hindered ketone and potential epimerization were not a problem.

[00:18:47]To recap, this group is the masked, stabilized version of the cyclopentadiene we need for the extreme Diels-Alder reaction.

[00:18:54]As a second step, the silyl protecting groups were removed by controlled treatment with TBAF.

[00:18:59]Note that this does not touch the triethyl silyl ethers as we will see at the end, these groups needed some more forcing conditions.

[00:19:07]Next up, we have two steps in one oxidation of the primary alcohol to the acid with TEMPO and hypervalent iodine led after stirring for 36 hours to the lactone.

[00:19:19]Finally, the elimination of the beta-hydroxy group via mesylation and heating gave the dienophile for our Diels-Alder reaction.

[00:19:27]Even though the upcoming reaction is pretty complex, the setup was pretty simple.

[00:19:32]The final conditions check out the supporting information for more details were 6-day long small-scale reactions with a sub-stoichiometric amount of a tetrazine trigger, giving the best yield based on recovered starting material.

[00:19:45]But what s the mechanism?

[00:19:47]First, we have the unmasking of the cyclopentadiene.

[00:19:50]This works through an inverse-electron demand Diels-Alder addition of tetrazine to norbornadiene, and subsequent nitrogen release through a retro-Diels-Alder this provides for a strong driving force.

[00:20:03]And as DJ Khaled would say, a second retro-Diels-Alder kicks out an aromatic pyridazine, fully de-masking the cyclopentadiene.

[00:20:14]Again, I recommend the nerds to read the SI as they nicely describe that the reactivity of tetrazines had to be well balanced electron-poor tetrazines reacted rapidly in the first step, but failed to ultimately fragment and release the pyridazine and cyclopentadiene.

[00:20:31]After uncovering the beast, the fourth Diels-Alder reaction with the unsaturated lactone gives the product which we initially envisioned in our retrosynthetic analysis.

[00:20:41]You may wonder why would we opt for sub stoichiometric amounts of trigger?

[00:20:45]The final product has a double bond and guess what this thing just doesn t want to stop Diels-Aldering.

[00:20:52]Too much free tetrazine left over leads to a competitive decomposition of the product.

[00:20:56]Alright, let s finish this thing.

[00:20:59]You probably remember we want an alpha-methyl ketone instead of the alkene.

[00:21:02]We can get there by an epoxidation and subsequent methyl addition with a Gilman reagent.

[00:21:08]The regioselectivity was thankfully controlled by disfavored 1,3-diaxial interactions between the methyl groups.

[00:21:16]The methyl nucleophile adds to the more distant, desired position and does so from the bottom face where its trajectory is not blocked by the carbon bridge.

[00:21:25]If you think this seems simple, don t fool yourself very precise reagent control was needed to not mess things up.

[00:21:32]The resulting alcohol was oxidized to the ketone present in the natural product, and the protecting groups were fluorinated away.

[00:21:39]Now, it s just two different esters groups to be introduced at the alcohols.

[00:21:43]Interestingly, the isopropyl ester proved less reactive as reaction with 11 equivalents of activated agent led to 43% of tertiary esterification and 26% of secondary esterification.

[00:21:56]Last, another large dump of anhydride and careful purification from excess reagents delivered the natural product.

[00:22:04]What a ride, we started with diving for sponges and ended in hardcore organic chemistry.

[00:22:10]I must say I m personally a bit diene myself after compiling this video.

[00:22:14]I hope you liked it and were to follow or at least learn some new things, regardless of your level of synthesis expertise.

[00:22:22]Once again, thank you for watching and supporting my videos!

[00:22:25]As always, until next time!