Chapter I of this thesis concentrates on new developments in the area of oxa-[3 +3] cycloaddition reaction, in particular, Lewis acid catalyzed version of this reaction.Synthetic scope a
Formal [3 + 3] Cycloaddition Reaction Catalyzed by
IntrOductiOn - ch nh nh kg 1
Cycloaddition and annulation reactions remain a widely used method for the construction of cyclic compounds An attractive feature, which makes these reactions a powerful synthetic tool, is their ability to conveniently build multiple bonds simultaneously with regio- and stereochemical control leading to polycyclic carbocycles and heterocycles For several years in our laboratory we have been investigating a formal
[3 + 3] cycloaddition reaction.' During this time we have demonstrated the versatility of the [3 + 3] cycloaddition reaction and also applied it to syntheses of several natural products.ˆ“
O AcO Xã Oo NR, © ae | C-1,2-addition ị
O Ôn lọ) electrocyclic bề ring-closure Oo
Mechanistically, this reaction proceeds through the sequence shown in Scheme 1.
This is a tandem process that involves a step-wise Knoevenagel-type condensation between an œ,B-unsaturated iminium salt 1 and a diketone equivalent 2 This process consists of C-1,2-addition followed by B-elimination 6n-electron electocyclic ring- closure of 1-oxatriene 3 leads to the formation of 1-oxadecalin 4 The result of this process is the formation of two new o-bonds in addition to generation of a new stereocenter adjacent to the heteroatom It can be considered formally an equivalent of a
[3 + 3] cycloaddition in which the three carbon atoms of œ,B—unsaturated iminium salt 1 have been added to the two carbon atoms and one oxygen atom of diketone The term [3
+ 3] cycloaddition was adapted by us from Seebach’s work describing a Stork-type carbo-[3 + 3] annulation reaction between nitroalkenes and enamines.” Two types of hetero-[3 + 3] cycloadditions have been studied in our lab and we termed them aza-[3 + 3] and oxa-[3 + 3] which produce l-azadecalin and 1-oxadecalin respectfully These reactions can be further classified as intermolecular or intramolecular.
This Chapter gives a brief history of formal [3 + 3] cycloaddition reaction by providing summary of previous work in our, as well as, other research laboratories but predominantly focuses on new developments in the area of [3 + 3] cycloaddition In particular, it comprehensively describes our effort to develop a new Lewis acid catalyzed method and its advantages, as well as, limitations Detailed experimental procedure is included at the end Selected 'H NMR spectra could be found in the Appendix.
The [3 + 3] cycloaddition reaction was pioneered by Link in 1944 who demonstrated that 4-hydroxycoumarin 5 can be condensed with a number of a,f- unsaturated ketones 6 to give [3 + 3] type products (Scheme 2).° Michael type additions were carried out by refluxing 5 and 6 in pyridine and the condensation products 7 were converted to the corresponding cyclic methyl ketals by refluxing in methanolic HCl, followed by elimination to provide 8 This transformation required harsh conditions; in addition, the reaction required an extra elimination step to produce a [3 + 3] type product.
` Ri pyr, reflux OH R' _1) MeOH, HO MeOH, HCl Ọ en PM LA
Co R2 _AoO.HOO, Ac;O, HCIO¿ oO” *O 7
The next step was taken several years later by de Groot, who studied the reaction between cyclic 1,3-diketone and an œ,B-unsaturated aldehyde as illustrates in Scheme 3.’
When dicarbonyl 9 and aldehyde 10 were refluxed in dry pyridine, high yields of the cyclized product 11 resulting from 1,2 addition were obtained The use of enals proved to be crucial for the further development; it gave the reaction a new twist and also demonstrated the potential use of the [3 + 3] cycloaddition It was now a one-pot reaction capable of making oxadecalins, such as 11, in moderate to high yields Unfortunately, we struggled to repeat these results in our laboratories An important observation was also made The condensation product 12, resulting from an acyclic dione 13 and enal 14, existed in both closed and open form Compounds 12 and 15 could be separated, but both gradually equilibrated to the original mixture. pyridine ẹ
Another maJor development came from the detailed study by Moreno-Mafias in
1985.° He disclosed the reaction that involved condensation of 6-methyl-4-hydroxy-2- pyrones 16 with œ,/đunsaturated aldehyde catalyzed by piperidinium acetate leading to a variety of products 17-21 (Scheme 4) These compounds, resulting from various competing pathways, were scrupulously isolated and analyzed When crotyl aldehyde (14) was used, the synthetically most useful [3 + 3] product 17 was isolated, unfortunately in a low yield Despite the synthetic potential of this formal [3 + 3] cycloaddition reaction, applications have been limited due to competing reaction pathways resulting from 1,2- versus 1,4-addition, as well as, C-addition versus O- addition.
Our group became interested in this reaction mainly because of its potential application to total synthesis of arisugacin A At about the same time that our group began to investigate this reaction, the Hua group reported an elegant study using cyclic enals, such as 22 and 23 (Scheme 5).” A major contribution was made by employing L- prolin as a suitable catalyst, as it afforded higher yields of the desired cycloadducts 24 and 25. ĩ O fe) 0.5 eqL-proline_
Formation of cycloadduct 25 as a single isomer hinted that this reaction is probable reversible as a more thermodynamically favorable isomer was formed This also introduced a problem of developing an asymmetric version of [3 + 3] cycloaddition.
Induction of a new stereogenic center would be problematic to accomplish stereoselectively if the reaction is reversible and oxatriene is involved, as it would equilibrate to 1:1 mixture.
Guided by these previous studies, our group was eager to develop a more general methodology that would allow the use of acyclic enals and also solve the problem of competing reaction pathways The use of pre-generated ơ,B-unsaturated iminium salts proved to be general and efficient solution for the regioselectivity problem This innovative development allowed our group to explore this useful formal [3 + 3] cycloaddition in depth and apply the new methodology to syntheses of several natural products (Figure 1).
CO,H rhododaurichromanic acid A/B ô9 4a:B0-clopi pumiiotoxin oon
Figure 1 Some completed natural products.
Still there remain two significant problems High temperatures are required (80 °C
- 150 °C) in most cases and pre-generation of iminium intermediates is necessary prior to the addition of diketone to control the regiochemistry and enhance the overall yield This called for a more general methodology which is simple and proceeds under milder reaction conditions Exploring the use of Lewis acids was an answer.
Lewis Acids and Formal [3 + 3] Cycloaddition
As our efforts of employing several Lewis acids as catalysts in our formal [3 + 3] cycloaddition reaction were in a full swing, we discovered that there were two reports in the literature that touched this issue In particular, Lee reported synthesis of 2H-pyrans by indium (III) chloride-catalyzed reactions of 1,3-dicarbonyl compounds with a variety of œ,B-unsaturated aldehydes in moderate yields (Scheme 6).'° Reaction of diketone 9 with crotyl aldehyde in refluxing acetonitrile for 4h in the presence of 50 mol % of indtum(IID) chloride afforded compound 11 in a 70% yield Reactions of 4-hydroxycoumarin 5 and 3- hydroxy-1H- phenalen-1-one 26 with several enals were also examined Unfortunately, authors only examined InCl; and reaction conditions were harsh to make it a versatile synthetic tool In addition, cyloadducts were generally obtained in moderate yields.
QInGls CHạCN, OH o OO reflux, 4h 11: 10% OY
OH ° Oo @® reflux, 4h @ Sr | bề ——
There is a brief mention of Lewis acid as a catalyst for [3 + 3] cycloaddition reaction in Shaw’s work (Scheme 7).'' One of the results they report is an annulation reaction between 4-hydroxycoumarin 5 with enantiopure 2,3-dideoxy-œ,B-unsaturated carbohydrate enal 29 in the presence of TiCl, in acetonitrile at 0 °C which proceeded in
46% to give a 1:1 mixture of diastereomers 30 As in the previous case, there was no effort made to optimize reaction condition or to attempt the use of a different Lewis acid.
In addition, better yield were obtained when L-prolin was used as a catalyst.
NN : O CH,CN, 6h œŒ*o OBn 46%
Although we should not be exactly considered pioneers in a field of Lewis acid catalyzed [3 + 3] cycloaddition reaction, it is obvious that not a lot of work has been done in this area and a more extensive study was needed that would address issues with low yield, harsh reaction conditions, and resourcefulness of this reaction Moreover, we stumbled upon the above reports when our own project had been well under way and a large amount of data collected.
Initial Developments c cớ 9
A detailed account of our Lewis acid catalyzed formal [3 + 3] cycloaddition reaction is discussed below As we were searching for a versatile reaction conditions for this reaction, two things we needed to improve: make reaction milder and eliminate an extra step, such as making iminium salt prior to addition of diketone.
Scheme 8 illustrates our first attempt with a Lewis acid, BF3-OEt, as a catalyst for a [3 + 3} cycloaddition reaction between enal 31 and a simple cyclic diketone 32.”
Desired cycloadduct 33 was obtained in a very high 95% yield In fact, no other products were detected by NMR More importantly, the procedure was simplified compared to our classical [3 + 3] cycloaddition Only standard aqueous work-up was needed; to neutralization of BF3;-OEt, with sat aq NaHCO; was followed by drying Compound 33 was not purified further It was very pure, by NMR and LCMS and was collected as colorless oil, without any yellowish tint, as it is usually a case with classical [3 + 3] cycloadducts, even after flash column chromatography and activated charcoal treatment.
Practically, the reaction was very easy to set up It simply involved mixing enal
31 with diketone 32 in a dry CHạC]; followed by addition of BE:-OEt; 4 A molecular sieves were needed to absorb water that formed during the course of the reaction.
Significantly, elevated temperature was needed to promote this reaction.
To find optimal conditions and to investigate the scope of our new methodology, we decided to initially study a temperature effect Lower temperatures are essential to minimize equilibration in order to have any chance for asymmetric version of [3 + 3] cycloaddition The same reaction between 31 and 32 as shown in Scheme 8 proceeded smoothly at 0 °C to provide the desired cycloadduct (Table 1, entry 2) Unfortunately, decrease in yield, 86% from 95%, was observed This could be attributed to a slower reaction, as some of the starting material, enal 31 was seen in the crude 'H NMR.
Although, when excess of enal 31 was used and the reaction was allowed to stir longer at
0 °C, no improve in yield was seen, possible due to competing reaction pathways Lower temperature, -10 °C, (entry 3, Table 1) was also attempted In this case, reaction never went to completion. entry BF3-OEt: (equiv) Time (h) Temp (0°C) yield of 33
4 0.1 16 23 80 50.05 16 23 65 Table 1 Temperature and Stoichiometry Studies.
To probe if the reaction between enal 31 and diketone 32 was indeed catalytic, we attempted it with 10 % and 1 % of the catalyst, BF3-OEt:, entries 4 and 5 in Table 1.
Results were impressive Although the yield dropped down to 65% when 1% of catalyst was used, it demonstrated the robust nature of this reaction The latter discovery opened doors for the more expensive catalysts Experimentations with different solvents, extreme addition did not improve the original protocol We were now ready to set on a quest to determine a scope of the useful Lewis acid.
A preliminary screening revealed that a range of Lewis acids could be used to promote this formal cycloaddition as summarized in Table 2 This was truly a screening process and no great care was taken to improve the reaction yields by optimizing reaction conditions or simply by repeating them In addition, 10% of Lewis acid was used for screening purposes for two reasons: some of the catalysts are expensive and to show how robust the reaction is Again, results were impressive Any Lewis acid we could lay our hands on catalyzed the cycloaddition reaction. entry LA equiv Time(h) Yield of 33
Table 2 Screening for Lewis acids to promote [3 + 3] cycloaddition reaction.
Few words have to be said about our preliminary screening In case of CuCl, a trace amount of by-product 34 was seen in 'H NMR It is believed to arise from a competing C-1,4 addition as illustrated in Scheme 9 We are not completely sure about the structural assignment of compound 34, as it was only seen in crude 'H NMR and was never isolated in a pure form Traces of 34 were also detected when InCl; and ZnBr2 were used Another important practical observation was made In case of triflate salts,
Sn(OTf),, Mg(OTf)2, Zn(OTf)2, and In(OTf)2, work-up was somewhat complicated as extensive foam formation was seen during aqueous work-up, resulting in lower yields.
A ot Oh; 1 és CuCl CỐ O
0” “OH Ct Aad 4-add C-1,2-add | 0
Remarkably, our Lewis acid catalyzed [3 + 3] cycloaddition reaction was promoted by even milder Lewis acids such as CuCl and ZnBr, and again proved to be truly catalytic when 1 % of TiCl, or ZnBr2 was used After laying down the foundation, we were now ready to explore the scope of this newly developed methodology.
Initially, our group started a strong research program in the area of [3 + 3] cycloaddition chemistry with the reaction of 6-methyl-4-hydroxy-2-pyrone with various enals, so as our efforts to develop a Lewis acid catalyzed version progressed, we decided to go back to the roots and revisit the above reaction Table 3 summarizes our discoveries. lo ọ lo
46 31 35 entry Lewis acid equiv conc [M] time (h) yield
9 ZnBr; 0.10 0.09 24 48 Table 3 Screening of Lewis Acids for the Reaction of 6-methyl-4-hydroxy-2-pyrone.
A problem of solubility was encountered during this study, as a result reactions were run at a lower concentration In the case of BF3-OEt, (entry 1-4), yield dropped dramatically when less catalyst was used Again, several Lewis acids were useful for this reaction, however in case of AICI; and In(OTf)2, yield were very low This could be due to a strong nature of AlCl; Unfortunately, these low yielding reactions were too messy to rigorously analyze and the only isolatable product was cycloadduct 35 No further optimization was done, since we were too impatient to move forward Hopefully, this reaction will be revisited in the future and at least a different solvent system would be tried. x® ° © fe) NHR>
Xe Coke XS OH -"" ÀR O R O 1 1
O 6n ie) electrocyclic bề ring-closure a
Scheme 10 illustrates our proposed mechanism and also shows similarities between the classical conditions and Lewis acid catalyzed process We believe that a complex between Lewis acid, diketone and enal (36) forms first, followed by probable C-
1,2-addition to give the same intermediate 37, when iminium salt is involved If that holds true, than the rest of the reaction, most likely, proceeds via the same mechanism as our classical [3 + 3] cycloaddition At this point, we are not sure about the true nature of the reaction mechanism and the big question still remains about the involvement of the
Lewis acid at the later stages of the reaction pathway, as it could prove to be crucial in developing asymmetric version of this reaction In addition, if complex between Lewis acid and reactant exists, it might be possible to control the regiochemical outcome These questions will be addressed later in this chapter.
After all of the above experimentation, we discovered that Lewis acids did not alter chemistry and provided the same 2H-pyran as when using iminium salts In addition, Lewis acids could be employed at sub-stoichiometric amounts, reaction temperatures could be lowered to room temperature, and BF3-Et,O and TiCl, became favorite Lewis acids because of their respective reactions were faster than those that employed weaker Lewis acids such as zinc and copper halides, and In(OTf)3 Being successful in our initial beginnings, we proceeded to explore the scope of our newly developed methodology to determine its synthetic value.
The Scope and Synthetic Value cà 15
Protective groups are useful tactical tools in organic synthesis but they play a passive role and their removal adds an extra step, which can be a nuisance during a multi- step, as well as, short synthesis These synthetic transformations are, in many cases, trivial but still require time and effort to carry them out successfully It is greatly desirable to minimize such operations.
As we moved forward with our Lewis acid catalyzed [3 + 3] cycloaddition, we decided to tackle this problem by exploring the possibility of in situ deprotection of diketone equivalents, followed by a consecutive cycloaddition reaction Importantly, our success in employing Lewis acids in promoting cycloadditions of vinylogous esters would be viewed as a major advantage of this methodology Some of our findings are summarized in Table 4.
One 1z CO OR O O 39 entry R= Lewis wail equiv Temp Time yield of 39
Table 4 Examples of Lewis acid catalyzed [3 + 3] cycloaddition of vinylogous ester.
As shown in Table 4, TIPS vinylogous ester 42 appears to work the best, leading to pyran 39 in 95% yield and 1.0 equiv of TiCl, was preferred, or the reaction was rather slow (entry 5) The reaction of methyl vinylogous ester 40 was very slow (entry 1), whereas TMS vinylogous ester 41 was feasible In general, all the attempted reactions of protected vinylogous esters were slower and lower yielding than those of their equivalent, free diketones This could be, in part, due to the decreased stability of the product 39 as it has a tertiary carbon adjacent to oxygen instead of quaternary and with longer reaction times it could decompose easier Entries 6 and 7 are important, as it showed that JacobsenCo-salen catalyst can be used to promote the cycloaddition This catalyst could potentially make this reaction asymmetric The above findings are significant because it allows one to protect diketones as vinylogous esters while carrying out other transformations, and subsequently, pursue the formal cycloaddition without the concern of having to hydrolyze vinylogous esters back to dikeones.**
To understand more about the mechanism of this reaction and to expand its use, we decide to try some diastereoselective cases It was previously shown in our group that when chiral enals are used, [3 + 3] cycloaddition reaction via iminium salt is diastereoselective Thermodynamic product as expected was winning the battle, as the reaction proved to be reversible Using chiral enals (S)-(-)-perillaldehyde (23) and (R)-(-
)-myrtenal (43), Lewis acid promoted formal [3 + 3] cycloadditions with pyrone 16 and diketone 32 gave the respective pyrans 25 and 44, and 45 and 46 in good yields as well as high diastereoselectivity (Scheme 11).
The major diastereomers in the above cases appear to be the more stable isomers, with AE ranging from 0.40 to 1.11 kcal mol' The stereochemical assignment relied on nOe The ratios of diastereomers were assigned based on crude 'H NMR In addition,
TiCl, proved to be the best catalyst for this reaction, as BF3;-OEt, and ZnBr; catalyzed reactions, were sluggish and low yielding Above results demonstrated that pyrone 16 does not necessarily gives products in low yield as cycloadducts 25 and 44 were obtained in 61% and 80% respectfully.
Very useful and practical development came when we were able to effectively carry out the reaction of Funk's enal 47” with pyrone 16 and diketone 32 to give the desired pyrans 48 and 49 in 71% and 77% yields, respectively However, rt was not sufficient enough to promote this reaction and alternatively 120 °C was need The reaction was allowed to run for 24 h in a sealed tube Although the reaction temperature was quite high this time, it has been very difficult to use a-oxygenated enals in this annulation reaction under the iminium salt conditions despite the potential of 48 and 49 in synthetic applications. on A OTBS ZA O
O TiCl, (0.12 equiv) T¡CH¿ (0.12 equiv) SS O
A more interesting finding, form a stand point of reaction mechanism, is demonstrated in Scheme 13 When ZnBr; catalyzed reaction between citral 50 and diketone 32 was allowed to react at 0 °C for 5 d, compound 51 was obtained in 78% yield It is believed to arise from a tandem [3 + 3] cycloaddition to give cycloadduct 52, an alkene cyclization, as carbocation 53 is trapped by a hydroxyl group This interesting result came as a complete surprise to us, as we were not planning to perform this tandem reaction sequence but were trying to produce more intermediate 52 for our parallel project, synthesis of naturally occurring chromanoids (Chapter II) Needless to say, our feelings were mixed but the fascinating chemistry overweight the frustration In addition, this reaction was performed on 8 g scale, based on citral We tried to achieve the similar result by using enal with a longer prenyl chain, in order to promote [3 + 3] cycloaddition followed by a polyene cyclization We did not succeed and the only isolatable product was compound 59 (Figure 2).
As illustrated in Figure 2, our protocol proved to be general as we prepared more pyrans 11, 17, 52, 54-59.
Figure 2 The Scope of Lewis Acid Catalyzed [3 + 3] Cycloaddition
The range of a, B-unsaturated aldehydes was used, as well as, different diketone equivalents To obtain compounds 11, 17, 54-56 and 58, BF3-OEt, was used as a catalyst In the case of 52 and 57, 0.5 equiv of TiCl, proved to work the best All the above reactions were run in anhy CHạC]; with 4 A molecular sieves Reaction times usually varied between 16 to 24 h Importantly, when hydroxy pyrone 16 was used cycloadducts 17, 54 and 55 were obtained in a much better yield 53%, 59% and 53% respectfully, compared to a simpler system shown in Table 3 As mentioned above, there was no cyclized product 51 detected, when citral (50) was used and compound 52 was the sole product Interestingly, when BF;-OEt;ạ was used to promote the same reaction, decomposition of the starting material was seen quickly by TLC and all of the starting material was gone in the first 30 min.
Compound 59 was only accessible when 0.5 equiv of Zn(TfO)¿ was used (Figure
2) The reaction was slow at 0 °C and it took 5 d for it to go to completion, since higher temperatures proved deadly When other stronger Lewis acids were used there was no product or starting material detected, as the reaction decomposed very quickly according to TLC When we attempted this reaction, we were hoping for the tandem process that would involve polyene cyclization but no such product was detected.
The above results demonstrate a wide scope of our Lewis acid catalyzed [3 + 3] methodology Two most practical findings that came out of this study are the use of protected vinylogous esters, since it eliminates the extra deprotection step and employment of a-oxygenated enals to provide useful synthetic products.
To understand more about the mechanism of the reaction and to probe our protocol for more synthetically useful applications we decided to investigate the regiochemical control that certain Lewis acids might have on the [3 + 3] cycloaddition.
It was previously shown in our group that when the pre-generated iminium salt of enal 31 is reacted with diketone 60 a mixture of regioisomers 61 and 62 arises (Scheme
14).'° Since our classical [3 + 3] cycloaddition is controlled by thermodynamics via reversibility of the ring closure, predominantly stable, in the above case 61, is formed in preference Otherwise one would expect to get roughly 1 : | ration of products 61 and 62.
If a method to control regioselectivity existed, it would prove to be invaluable in cases when the formation of regioisomers is possible.
We decide to study the same model system as shown in Scheme 14 Two isomers
61 and 62 were readily separateable on the TLC plate or by flash silica gel chromatography However, spectroscopically they were almost identical except for the methyl resonances at 1.10 and 1.9 ppm respectfully Our preliminary calculations using
PM3 level of theory and SPARTAN software indicated that the more thermodynamically favored isomer is 61, which was expected from the previous results Our first attempt with BF3-OEt, as a catalyst is shown in Scheme 14.
1:1 mixture DEM, 48h," 48h, rt of 67 and 62
Overall Conclusion HH nh kh nh sy 32
In conclusion, Lewis acid catalyzed [3 + 3] cycloaddition reaction was described. This methodology, in most cases, is superior to our classical protocol It features the ease of handling, ambient temperature, easy work-up, and, simplified purification It also proved to be quiet versatile, as a range of Lewis acids, as well as, different enals and diketones were investigated In general, this new protocol was found to give similar results, as one with iminium salts However, it exhibited an interesting preference for the kinetic product, although this preliminary result has to be definitely explored further A biggest synthetic advantage of this methodology is its ability to promote the cycloaddition reaction between an enal and a protected diketone, since it conveniently eliminates an extra deprotection step Another important result is the use of a-oxygenated enal to provide a useful synthetic intermediate Unfortunately, some limitations still exist.
If this methodology would be explored further, the immediate goal would be to develop an asymmetric version of this formal [3 + 3] cycloaddition
(a) Hsung, R P.; Shen, H C.; Douglas, C J.; Morgan, C D.; Degen, § J.; Yao, L J J Org. Chem 1999, 64, 690 (b) Hsung, R P.; Wei, L.-L.; Sklenicka, H M.; Douglas, C J.; McLaughlin,
M J.; Mulder, J A.; Yao, L J Org Lett 1999, 1, 509 (c) Shen, H C.; Wang, J.; Cole, K P.; McLaughlin, M J.; Morgan, C D.; Douglas, C J.; Hsung, R P.; Coverdale, H A.; Gerasyuto, A. I.; Hahn, J M.; Liu, J.; Wei, L.-L.; Sklenicka, H M.; Zehnder, L R.; Zificsak, C A J Org. Chem 2003, 68,1729.
For pyranyl heterocycle construction and related natural product syntheses, see: (a) Hsung, R P.; Cole, K P.; Zehnder, L R.; Wang, J.; Wei, L L.; Yang, X.-F.; Coverdale, H A Tetrahedron
2003, 59, 311 (b) Cole, K P.; Hsung, R P.; Yang, X.-F Tetrahedron Lett 2002, 43, 3341 (c) Wang, J.; Cole, K P.; Wei, L L.; Zehnder, L R.; Hsung, R P Tetrahedron Lett 2002, 43, 3337. (d) Cole, K P.; Hsung, R P Tetrahedron Lett 2002, 43, 8791 (e) McLaughlin, M J.; Hsung, R.
P J Org Chem 2001, 66, 1049 (f) McLaughlin, M J; Shen, H C.; Hsung, R P Tetrahedron Lett 2001, 42, 609 (g) Zehnder, L R.; Wei, L.-L.; Hsung, R P.; Cole, K P.; McLaughlin, M J.; Shen, H C.; Sklenicka, H M.; Wang, J.; Zificsak, C A Org Lett 2001, 3, 2141 (h) Zehnder, L. R.; Hsung, R P.; Wang, J.; Golding, G M Angew Chem., Int Ed 2000, 39, 3876 (1) Zehnder,
For syntheses of oxygen containing heterocycle see: (a) Cole, K P.; Hsung, R P Org Lett 2003,
5, 4843 (b) Kurdyumov, A V.; Hsung, R P.; Ihlen, K.; Wang, J Organic Lett 2003, 5, 3935. For review see: (a) Harrity, J P A.; Provoost, O Org Biomol Chem 2005, 3, 1349 (b) Hsung,
R P.; Kurdyumov, A V.; Sydorenko, N Zur J Org Chem 2005, 23-44.
Seebach, D.; Missbach, M.; Calderari, G.; Eberle, M J Am Chem Soc 1990, 112, 7625-7638. (a) Ikawa, M.; Stahmann, M A.; Link, K P J Am Chem Soc 1944, 66, 902-906 (b) Seidman, M.; Robertson, D N.; Link, K P J Am Chem Soc 1950, 72, 5193-5195 (c) Seidman, M; Link,
K P J Am Chem Soc 1952, 74, 1885-1886. de Groot, A.; Jansen, B J M Tetrahedron Lett 1975, 39, 3407-3410. de March, P.; Moreno-Mafias, M.; Casado, J.; Pleixats, R.; Roca, J L.; Trius, A J Heterocycl. Chem 1984, 21, 85-89.
Hua, D H.; Chen, Y.; Sin, H.-S.; Maroto, M J.; Robinson, R D.; Newell, S W.; Perchellet, E. M.; Ladesich, J B.; Freeman, J A.; Perchellet, J P.; Chiang, P K J Org Chem 1997, 62, 6888- 6896.
Lee, Y R.; Kim, D H.; Shim, J.-J.; Kim, S K.; park, J H.; Cha, J S.; Lee, C.-S Bull Korean.Chem Soc 2002, 23, 998.
Sagar, R.; Singh, P.; Kumar, R.; Maulik, P R.; Shaw, A K Carbohydr Res 2005, 340, 1287. Kurdyumov, A V.; Lin, N.; Hsung, R P.; Gullickson, G C.; Cole, K P.; Sydorenko, N.; Swidorski, J J Org Lett 2006, 8, 191-193 ;
For the synthesis of aldehyde 47, see: Aungst, R A Jr.; Funk, R L Org Lett 2001, 3, 3553. Cossy, J.; Rakotoarisoa, H.; Kahn, P.; Desmurs, J -R Tetrahedron Lett 1998, 39, 9671.
Y Kang, Y Mei, Y Du, Z Jin, Org Lett 2003, 5, 4481-4484.
Pettigrew, J D.; Cadieux, J A.; So, 8 S 5; Wilson, P D Org Lett 2005, 7, 467.
Gerasyuto, A.L; Hsung, R.P.; Sydorenko, N.; Slafer, B J Org Chem 2005, 70, 4249
Column chromatography was performed on Bodman silica gel (60 A, 230-400 mesh). THF was distilled over sodium/benzophenone under nitrogen Dichloromethane, toluene, acetonitrile, diisopropyl amine, triethylamine, and benzene were distilled from calcium hydride under nitrogen Methanol was distilled from magnesium and stored under nitrogen over 3 A molecular sieves N,N-dimethylformamide was distilled under vacuum from calcium hydride and stored over 4 A molecular sieves Flasks were flame dried under vacuum and purged with nitrogen before use TLC plates were silica (Whatman, polyester backed) and were visualized with UV (254 nm) and either anisaldehyde or permanganate stains IR spectra were recorded on NaCl plates using a Midac M2000 FTIR IR intensities are reported as follows: w = weak, m = medium, s = strong, vs very strong, br = broad 500 MHz 'H spectra were recorded on a Varian Inova spectrometer; 300 MHz spectra were recorded on a Varian Unity or Varian Inova instruments and are referenced to TMS at 5 0.00 ppm Other NMR solvents are referenced to residual solvent NMR peak multiplicities are reported as follows: s singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br = broad ‘°C spectra were recorded on a Varian Inova spectrometers at 125 and 75 MHz and are referenced to the center chloroform peak at 5 77.23 ppm Electrospray mass spectra were recorded on a
Bruker Biotof II ESI-TOF/MS using either PPG or PEG standards as high resolution calibrants Unless noted, all reagents (Acros, TCI, Aldrich) were used as received.
General procedure for annulation of enals with 1,3-diketones catalyzed by a strong Lewis acid The following reaction procedure is representative:
Compound 33: To the solution of aldehyde 31 (50 mg, 0.59 mmol) in anhydrous CH2Ch (1.5 mL) was added 1,3-diketone 32 (66 mg, 0.59 mmol) and 4 A MS (45 mg) followed by the dropwise addition of 1.0 M solution of TiCl, in CH2Cl, (0.06 mL, 0.059 mmol). The reaction mixture was stirred for 16 h at room temperature at which point it was poured into water (0.5 mL) The reaction was worked-up with sat aq NaHCO; until basic, washed with sat aq NaCl and dried over Na2SO, Evaporation of the solvent under the reduced pressure afforded 100 mg of 33 (95% yield).
General procedure for annulation of enals with 1,3-diketones catalyzed by a weak Lewis acid The following reaction procedure is representative:
ZnBr; (0.1 equiv), ọ lộ Be anhy DCM, 4AMS LO rt, 48h O 72%
Compound 33: To the solution of aldehyde 31 (50 mg, 0.59 mmol) in anhydrous CH2Ch
(1.5 mL) was added 1,3-diketone 32 (66 mg, 0.59 mmol) and 4 A MS (45 mg) followed by the addition of ZnBr; (13 mg, 0.059 mmol) The reaction mixture was stirred for 48 h worked-up with sat aq NaHCO; until basic, washed with sat aq NaCl and dried over
Na;SOa After the solvent was removed under the reduced pressure, the crude product was purified using flash silica gel column chromatography [20% EtOAc in hexanes] to give 76 mg of 33 (72% yield)
General procedure for annulation of enals with 4-hydroxy-2-pyrone The following reaction procedure is representative: fe) ọ oan
Compound 35: To the solution of aldehyde 31 (50 mg, 0.59 mmol) in anhydrous CHạC]; (4 mL) was added 4-hydroxy-2-pyrone 16 (74 mg, 0.59 mmol) and 4 A MS (90 mg) followed by the dropwise addition of 1.0 M solution of BF3Et,O in CH¿C]; (0.6 mL, 0.59 mmol) The reaction mixture was stirred for 24 h at room temperature at which point it was poured into water (1 mL) The reaction was worked-up with sat aq NaHCO; until basic, washed with sat aq NaCl and dried over Na2SO, After the solvent was removed under the reduced pressure, the crude product was purified using flash silica gel column chromatography [10% EtOAc in hexanes] to give 78.2 mg of 35 (69% yield)
General procedure for annulation of enals with vinylogous silyl esters The following reaction procedure is representative:
Compound 39: To the solution of aldehyde 38 (50 mg, 0.71 mmol) in anhydrous CHạC]; (1.5 mL) was added 42 (190 mg, 0.71 mmol) and 4 A MS (50 mg) followed by the dropwise addition of 1.0 M solution of T¡Cl¿ in CH2Ch (0.71 mL, 0.71 mmol) The reaction mixture was stirred for 24 h at room temperature at which point it was poured into water (0.5 mL) The reaction was worked-up with sat aq NaHCO; until basic, washed with sat aq NaCl and dried over Na;SOa After the solvent was removed under the reduced pressure, the crude product was purified using flash silica gel column chromatography [20% EtOAc in hexanes] to give 110 mg of 39 (95% yield).
Xi oo TiCl, (0.12 equi (0.12 say — O toluene 120 °C 48: 71%
Compound 48. via iminium salt To a flame-dried flask were added aldehyde 47 (0.04 mL, 0.2 mmol) and anhydrous EtOAc (4 mL) The solution was cooled to —10 °C, and piperidine (0.02 ml, 0.2 mmol) and then acetic anhydride (0.019 mL, 0.2 mmol) were added dropwise via syringe Reaction mixture was sealed under nitrogen and heated at 85 °C for | h The resulting iminium salt solution was transferred via a cannula to a suspension of pyrone 16
(0.025 g, 0.2 mmol) in anhydrous toluene (6 mL) in a flame-dried sealed tube The reaction mixture was sealed under nitrogen and heated at 120 °C in an oil bath for 24 h.
After the solvent was removed under the reduced pressure, the crude product was purified using flash silica gel column chromatography [25% EtOAc in hexanes] (deactivated with
1% EtạN) to give 29 mg of 48 (43% yield) as a pale yellow oil.
Lewis acid catalyzed To a flame-dried sealed tube were added 1g 4 A MS, 4-hydroxy-2- pyrone 16 (0.046 g, 0.25 mmol) and toluene (10 mL) The solution was stirred for 5 min and aldehyde 47 (0.05 mL, 0.25 mmol) and then TiCly (1M in CH2Ch, 0.04 mmol, 0.04 ml) were added dropwise via syringe The reaction mixture was sealed under nitrogen and heated at 120 °C for 24 h After the solvent was removed under the reduced pressure, the crude product was purified using flash silica gel column chromatography [25% EtOAc in hexanes] (deactivated with 1% EtạN) to give 54 mg of 48 (71% yield) as a pale yellow oil.
'H NMR (500 MHz, CDCl) 5 0.21 (s, 6H), 0.92 (s, 9H), 1.40 (d, 3H, J = 6.5 Hz), 2.19 (s, 3H), 4.81 (q, 1H, J = 6.5 Hz), 5.55 (s, 1H), 5.78 (s, 1H); °C NMR (75 MHz, CDC];) ồ
IR (film) cm 3054s, 2987s, 1704s, 1642m, 1424m, 838s; mass spectrum (APCI): m/e
Ry= 0.38 (20% EtOAc in hexanes); mp 127-130 °C; [a]p = -254.7 ° (c 0.63, CH;C];); 'H NMR (300 MHz, CDCIs) 8 0.89 (s, 3H), 1.33 (s, 3H), 1.43 (d, 1H, J = 10.5 Hz), 2.00—
135.5, 161.2, 162.7,164.7; IR (film) cm 2994m, 2361w, 1713s, 1636m, 1560m, 1448w; mass spectrum (APCI): m/e (% relative intensity) 259.3 (100) (M + H)’, 258 (51), 215 (2), 541 (2) m/e calcd for C16H1903 (M + H)* 259.1334, found 259.1502.
Ry= 0.28 (20% EtOAc in hexanes); [œ]p”°= -25.8 ° (c 0.57, CHạC];);
!H NMR (300 MHz, CDC];) 8 1.26 (ddd, 1H, J = 3.9, 14.7, 16.2 Hz), 1.70 (s, 3H), 1.61—
4.90 (dd, 1H, J = 4.8, 11.4 Hz), 6.10 (s, 1H); !C NMR (75 MHz, CDC];) ử 20.6, 20.7,
(neat) cm' 2940m, 2866w, 1652s, 1608s, 1403s; mass spectrum (APCI): m/e (% relative intensity) 245.1 (100) (M + H)*, 138 (12), 125 (31) m/e calcd for CigH2102 (M + H)*
Ry= 0.28 (20% EtOAc in hexanes); [œ]p””= -36.2 ° (c 0.55, CH;C];);
6.08 (d, 1H, J = 2.4 Hz); ''C NMR (75 MHz, CDC];) 5 20.9, 21.9, 25.2, 25.6, 27.9, 31.8, 35.4, 40.2, 41.9, 48.4, 72.0, 110.9, 114.5, 133.0, 172.2, 195.7; IR (film) cm 2948m,
2870m, 1654s, 1595s, 1400m; mass spectrum (APCI): m/e (% relative intensity) 245.1 (100) (M + H)*, 245 (26), 277 (21), 203 (16), 149 (12) m/e calcd for C16H2|02 (M + H)*
'H NMR (500 MHz, CDCl) 5 0.22 (s, 6H), 0.91 (s, 9H), 1.37 (d, 3H, J = 6.5 Hz), 1.92- 1.97 (m, 2H), 2.33-2.39 (m, 4H), 4.74 (q, 1H, J = 6.5 Hz), 5.62 (s, 1H); °C NMR (125
145.5, 165.9, 195.3; IR (film) cm! 3056s, 2986s, 1649s, 1611m, 1424s, 1396s, 845s; mass spectrum (APCI): m/e (% relative intensity) 295 (15) (M + H)’, 257 (10), 185 (60),
180 (15), 143 (100), 129 (10), 101 (10); m/e calcd for Ci¢H26O3Si (M + H)” 295.1729, found 295.1746
123.5, 131.9, 172.1, 194.9; IR (film) cm” 2969s, 2934s, 1723s, 1641s, 1591s; mass spectrum (ESI): m/e (% relative intensity) 247 (100) (M + H)”, 245 (10), 239 (5), 197 (5);m/e calcd for C16H2302 (M + H)" 247.1698, found 247.1658.
Experimental Procedures .: .c cà cv ii 35
Column chromatography was performed on Bodman silica gel (60 A, 230-400 mesh). THF was distilled over sodium/benzophenone under nitrogen Dichloromethane, toluene, acetonitrile, diisopropyl amine, triethylamine, and benzene were distilled from calcium hydride under nitrogen Methanol was distilled from magnesium and stored under nitrogen over 3 A molecular sieves N,N-dimethylformamide was distilled under vacuum from calcium hydride and stored over 4 A molecular sieves Flasks were flame dried under vacuum and purged with nitrogen before use TLC plates were silica (Whatman, polyester backed) and were visualized with UV (254 nm) and either anisaldehyde or permanganate stains IR spectra were recorded on NaCl plates using a Midac M2000 FTIR IR intensities are reported as follows: w = weak, m = medium, s = strong, vs very strong, br = broad 500 MHz 'H spectra were recorded on a Varian Inova spectrometer; 300 MHz spectra were recorded on a Varian Unity or Varian Inova instruments and are referenced to TMS at 5 0.00 ppm Other NMR solvents are referenced to residual solvent NMR peak multiplicities are reported as follows: s singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br = broad ‘°C spectra were recorded on a Varian Inova spectrometers at 125 and 75 MHz and are referenced to the center chloroform peak at 5 77.23 ppm Electrospray mass spectra were recorded on a
Bruker Biotof II ESI-TOF/MS using either PPG or PEG standards as high resolution calibrants Unless noted, all reagents (Acros, TCI, Aldrich) were used as received.
General procedure for annulation of enals with 1,3-diketones catalyzed by a strong Lewis acid The following reaction procedure is representative:
Compound 33: To the solution of aldehyde 31 (50 mg, 0.59 mmol) in anhydrous CH2Ch (1.5 mL) was added 1,3-diketone 32 (66 mg, 0.59 mmol) and 4 A MS (45 mg) followed by the dropwise addition of 1.0 M solution of TiCl, in CH2Cl, (0.06 mL, 0.059 mmol). The reaction mixture was stirred for 16 h at room temperature at which point it was poured into water (0.5 mL) The reaction was worked-up with sat aq NaHCO; until basic, washed with sat aq NaCl and dried over Na2SO, Evaporation of the solvent under the reduced pressure afforded 100 mg of 33 (95% yield).
General procedure for annulation of enals with 1,3-diketones catalyzed by a weak Lewis acid The following reaction procedure is representative:
ZnBr; (0.1 equiv), ọ lộ Be anhy DCM, 4AMS LO rt, 48h O 72%
Compound 33: To the solution of aldehyde 31 (50 mg, 0.59 mmol) in anhydrous CH2Ch
(1.5 mL) was added 1,3-diketone 32 (66 mg, 0.59 mmol) and 4 A MS (45 mg) followed by the addition of ZnBr; (13 mg, 0.059 mmol) The reaction mixture was stirred for 48 h worked-up with sat aq NaHCO; until basic, washed with sat aq NaCl and dried over
Na;SOa After the solvent was removed under the reduced pressure, the crude product was purified using flash silica gel column chromatography [20% EtOAc in hexanes] to give 76 mg of 33 (72% yield)
General procedure for annulation of enals with 4-hydroxy-2-pyrone The following reaction procedure is representative: fe) ọ oan
Compound 35: To the solution of aldehyde 31 (50 mg, 0.59 mmol) in anhydrous CHạC]; (4 mL) was added 4-hydroxy-2-pyrone 16 (74 mg, 0.59 mmol) and 4 A MS (90 mg) followed by the dropwise addition of 1.0 M solution of BF3Et,O in CH¿C]; (0.6 mL, 0.59 mmol) The reaction mixture was stirred for 24 h at room temperature at which point it was poured into water (1 mL) The reaction was worked-up with sat aq NaHCO; until basic, washed with sat aq NaCl and dried over Na2SO, After the solvent was removed under the reduced pressure, the crude product was purified using flash silica gel column chromatography [10% EtOAc in hexanes] to give 78.2 mg of 35 (69% yield)
General procedure for annulation of enals with vinylogous silyl esters The following reaction procedure is representative:
Compound 39: To the solution of aldehyde 38 (50 mg, 0.71 mmol) in anhydrous CHạC]; (1.5 mL) was added 42 (190 mg, 0.71 mmol) and 4 A MS (50 mg) followed by the dropwise addition of 1.0 M solution of T¡Cl¿ in CH2Ch (0.71 mL, 0.71 mmol) The reaction mixture was stirred for 24 h at room temperature at which point it was poured into water (0.5 mL) The reaction was worked-up with sat aq NaHCO; until basic, washed with sat aq NaCl and dried over Na;SOa After the solvent was removed under the reduced pressure, the crude product was purified using flash silica gel column chromatography [20% EtOAc in hexanes] to give 110 mg of 39 (95% yield).
Xi oo TiCl, (0.12 equi (0.12 say — O toluene 120 °C 48: 71%
Compound 48. via iminium salt To a flame-dried flask were added aldehyde 47 (0.04 mL, 0.2 mmol) and anhydrous EtOAc (4 mL) The solution was cooled to —10 °C, and piperidine (0.02 ml, 0.2 mmol) and then acetic anhydride (0.019 mL, 0.2 mmol) were added dropwise via syringe Reaction mixture was sealed under nitrogen and heated at 85 °C for | h The resulting iminium salt solution was transferred via a cannula to a suspension of pyrone 16
(0.025 g, 0.2 mmol) in anhydrous toluene (6 mL) in a flame-dried sealed tube The reaction mixture was sealed under nitrogen and heated at 120 °C in an oil bath for 24 h.
After the solvent was removed under the reduced pressure, the crude product was purified using flash silica gel column chromatography [25% EtOAc in hexanes] (deactivated with
1% EtạN) to give 29 mg of 48 (43% yield) as a pale yellow oil.
Lewis acid catalyzed To a flame-dried sealed tube were added 1g 4 A MS, 4-hydroxy-2- pyrone 16 (0.046 g, 0.25 mmol) and toluene (10 mL) The solution was stirred for 5 min and aldehyde 47 (0.05 mL, 0.25 mmol) and then TiCly (1M in CH2Ch, 0.04 mmol, 0.04 ml) were added dropwise via syringe The reaction mixture was sealed under nitrogen and heated at 120 °C for 24 h After the solvent was removed under the reduced pressure, the crude product was purified using flash silica gel column chromatography [25% EtOAc in hexanes] (deactivated with 1% EtạN) to give 54 mg of 48 (71% yield) as a pale yellow oil.
'H NMR (500 MHz, CDCl) 5 0.21 (s, 6H), 0.92 (s, 9H), 1.40 (d, 3H, J = 6.5 Hz), 2.19 (s, 3H), 4.81 (q, 1H, J = 6.5 Hz), 5.55 (s, 1H), 5.78 (s, 1H); °C NMR (75 MHz, CDC];) ồ
IR (film) cm 3054s, 2987s, 1704s, 1642m, 1424m, 838s; mass spectrum (APCI): m/e
Ry= 0.38 (20% EtOAc in hexanes); mp 127-130 °C; [a]p = -254.7 ° (c 0.63, CH;C];); 'H NMR (300 MHz, CDCIs) 8 0.89 (s, 3H), 1.33 (s, 3H), 1.43 (d, 1H, J = 10.5 Hz), 2.00—
135.5, 161.2, 162.7,164.7; IR (film) cm 2994m, 2361w, 1713s, 1636m, 1560m, 1448w; mass spectrum (APCI): m/e (% relative intensity) 259.3 (100) (M + H)’, 258 (51), 215 (2), 541 (2) m/e calcd for C16H1903 (M + H)* 259.1334, found 259.1502.
Ry= 0.28 (20% EtOAc in hexanes); [œ]p”°= -25.8 ° (c 0.57, CHạC];);
!H NMR (300 MHz, CDC];) 8 1.26 (ddd, 1H, J = 3.9, 14.7, 16.2 Hz), 1.70 (s, 3H), 1.61—
4.90 (dd, 1H, J = 4.8, 11.4 Hz), 6.10 (s, 1H); !C NMR (75 MHz, CDC];) ử 20.6, 20.7,
(neat) cm' 2940m, 2866w, 1652s, 1608s, 1403s; mass spectrum (APCI): m/e (% relative intensity) 245.1 (100) (M + H)*, 138 (12), 125 (31) m/e calcd for CigH2102 (M + H)*
Ry= 0.28 (20% EtOAc in hexanes); [œ]p””= -36.2 ° (c 0.55, CH;C];);
6.08 (d, 1H, J = 2.4 Hz); ''C NMR (75 MHz, CDC];) 5 20.9, 21.9, 25.2, 25.6, 27.9, 31.8, 35.4, 40.2, 41.9, 48.4, 72.0, 110.9, 114.5, 133.0, 172.2, 195.7; IR (film) cm 2948m,
2870m, 1654s, 1595s, 1400m; mass spectrum (APCI): m/e (% relative intensity) 245.1 (100) (M + H)*, 245 (26), 277 (21), 203 (16), 149 (12) m/e calcd for C16H2|02 (M + H)*
'H NMR (500 MHz, CDCl) 5 0.22 (s, 6H), 0.91 (s, 9H), 1.37 (d, 3H, J = 6.5 Hz), 1.92- 1.97 (m, 2H), 2.33-2.39 (m, 4H), 4.74 (q, 1H, J = 6.5 Hz), 5.62 (s, 1H); °C NMR (125
145.5, 165.9, 195.3; IR (film) cm! 3056s, 2986s, 1649s, 1611m, 1424s, 1396s, 845s; mass spectrum (APCI): m/e (% relative intensity) 295 (15) (M + H)’, 257 (10), 185 (60),
180 (15), 143 (100), 129 (10), 101 (10); m/e calcd for Ci¢H26O3Si (M + H)” 295.1729, found 295.1746
123.5, 131.9, 172.1, 194.9; IR (film) cm” 2969s, 2934s, 1723s, 1641s, 1591s; mass spectrum (ESI): m/e (% relative intensity) 247 (100) (M + H)”, 245 (10), 239 (5), 197 (5);m/e calcd for C16H2302 (M + H)" 247.1698, found 247.1658.
Syntheses of Chromenes and Chromanes
Biosynthetic Relationship .cccnee se 47
Based on the previously reported work on similar chromene systems and also guided by the fact that grifolin and orcinol has been isolated form rhododendron dauricum, the same plant daurichromenic acid was isolated, biosynthetic path could be proposed Goto and co-workers report conversion of grifolin to isogrifolin (16) by exposing 14 to HCI (Scheme 1).'”° Interestingly, isogrifolin possesses a chromane core
(no conjugated olefin) and not a chromene one as confluentin or daurichromenic acid. Reduron and co-workers isolated 7-demethylplastochromenol-2 (17) and a known sesquiterpene-substituted quinone (19), from the leaves of Seseli furreynii.'°? They proposed oxatriene 18 as an intermediated in the biosynthetic relationship between these two natural products.
NN ⁄ O_ keto-enol NN Z OH tautomerization
Proposed biosynthesis of daurichromenic acid is shown in Scheme 2 Although, orcinol was isolated from rhododendron dauricum together with grifolin and daurichromenic acid, farnesol, or its derivative, or any other compound that would explain the formation of the prenyl side chain, was never isolated from the same source as farnesylphenol type natural products Therefore proposed biosynthesis starts with grifolin.
Two pathways are possible The first on would involve the formation of dihydroconfluentin (20) followed by dehydration and carboxylation or formylation and oxidation sequence The second pathway involves oxatriene 21 which would produce confluentin by electrocyclic ring-closure Induction of the acid functionality seems more such compounds existed, iso-daurichromenic acid (iso-2), regioisomer of daurichromenic acid, might be formed Although, iso-2 could be isomarized to a more favored 2, where there are no steric interactions between chromene oxygen and carboxylic acid group, considering that a re-formation of the oxatriene ring is possible.
+ >> a >ằ oO electrocyclic OH ring-closure ``O : oO
The biosynthetic relationship between daurichromenic acid (2) and rhododaurichromanic acid A/B was realized by Kashiwada et al (Scheme 3).’ It has been shown that intramolecular photochemical [2 + 2] cycloaddition reaction of daurichromenic acid yields a mixture of rhododaurichromanic acid A and B. Isomerization of C11-C12 olefin geometry accompanies the [2 + 2] cycloaddition and is believed to occur prior to the photochemical ring closure, resulting in two cycloadducts la/b This proposed pathway is also supported by the fact that rhododaurichromanic acid
A and B were isolated in almost 1 : 1 ration, 75 mg and 69 mg, respectfully, as expected if they originate from daurichromenic acid via photochemical [2 + 2] cycloaddition via polyene cyclization of daurichromenic acid or its epoxy version 22 (Scheme 4) Mori and co-workers tried polyene cyclization in attempt to build hongoquercin B skeleton from the acyclic precursor but the reaction failed to provide the desired product.’
Unfortunately, authors did not elaborate further on the subject of the acyclic precursor used or reaction conditions. polyene polyene Ó €——— cyclization cyclization
Natural products from the same family as hongoquercin A and B, siccanin, as well as, closely related siccanochromene E are believed to come from a partially acyclic precursor 23 (Scheme 5) Nozoe and Suzuki reported the isolation of the triprenyl phenol derivatives presiccanochromenc acid (23), siccanochromenic acid (24) and siccanochromene A (25) from the same source as 6 and 7 4 Unfortunately, compound 26 was not found, which could be an ancestor of siccanin, siccanochromene E, and possibly hongoquercin A and B Nozoe and Suzuki also report the successful and quantitative conversion of methyl ester of presiccanochromenic acid to methyl ester of siccanochromenic acid by refluxing it in benzene in presence of DDQ Decarboxylation followed by cyclization and either internal trapping of the carbocation or elimination should provide siccanin or siccanochromene E, respectively Transformation of 23 to 24 could provide an insight as to how natural chromenes form.
23: presiccanochromenic 24: siccanochromenic 25: siccanochromene A acid ' acid internal elimination
T: siccanin 6: siccanochromene E 26: has not been reported
There will be more time devoted to the discussion of the polyene cyclization and its possible use in synthesis of both hongoquercin A and B Also an intriguing discovery which could suggestion the true biosynthetic pathway between daurichromenic acid and rhododaurichromanic acid A will be discussed in detail.
Previous Work: Rhododaurichromanic Acid A & B and
A range of biological activities of naturally occurring chromenes and chromanes attracted attention of several research groups It is important to develop a general and straightforward approach toward chromene or chromane core which could potentially be used to synthesize analogues of naturally occurring compounds There is also a great effort to develop an enantioselective methodology Several total syntheses of these natural products, as well as, some approaches are described in this section.'* The statement has to be made before discussing total syntheses of rhododaurichromanic acid
A and B and daurichromenic acid by other research groups We were the first once to accomplish total syntheses of these natural products ° and our work was in no way affected by any of the work described below Also this work is discussed because our own interest in this natural products, while other interesting and creative approaches toward different naturally occurring chromenes and chromanes are omitted.
A very elegant and straightforward approach to the chromene system was developed by Jin and co-workers.“ They used a unique microwave assisted tandem condensation and intramolecular SN;'-type cyclization reaction between resorcinol derivative 27 and a, -unsaturated aldehyde 28 in their synthesis of daurichromenic acid was found that in the absence of microwave irradiation, even at high temperatures and prolonged reaction times the reaction was very low yielding For example, under reaction conditions developed by Shigemasa,’’ Ca(OH)2, MeOH, reflux, 4 days, the reaction between 27 and 28 was extremely slow and low yielding, only 15% of the desired cycloadduct 29 was isolated Heating the mixture at 90 °C in a sealed reaction vessel for
1 d only improved the yield slightly After additional experimentation the optimized conditions were found in which the mixture of compound 27 (2.0 equiv), 28 (1.0 equiv)
CaC1;-2HO, NEt;, and EtOH was irradiated for 20 min The additional 1.0 equiv of compound 28 and 20 min of extra irradiation was needed to push the reaction to completion Using these optimized conditions allowed to isolate compound 29 in 70% yield Unfortunately, hydrolysis of ethyl ester 29 was carried out in poor yields due to the competing decarboxylation.
CaCIz-HzO, NEt; 3 M NaOH, EtO EtOH, microwave Eto’ > Oo ——— >- 2
Although this two-step synthesis is very concise and clever, there still were two aspects that authors decided to improve: low yield of saponification and the expensive starting material 27 To overcome these problems, resorcinol derivative 30 was chosen(Scheme 7) for the cycloaddition reaction Orcinol (15) was formylated using Vilsmeier-
Haack reaction to give the corresponding aldehyde!® which was sequentially oxidized to the carboxylic acid 31 Unfortunately, microwave assisted annulation of carboxylic acid
31 with enal 28 failed to provide the desired daurichromenic acid It was then decided to protect carboxylic acid 31 as a f-trimethylsilyl ethyl ester which could be easily converted back to the carboxylic acid Compound 31 was treated with 2-
(trimethylsilyl)ethanol under Mitsunobu conditions to give 30.'”
QH 9 HH TMSCH;CH,OH lu 1) POCl, DMF(98%) HQ DEAD, PPh, CH,Cly 2 2 › MS g
HạC OH DMSO, HạO (99%) HạC OH 90% HạC OH
To facilitate the cycloaddition reaction, 30 and 28 were mixed with CaCl,-H,O and NEt in EtOH in a sealed reaction vessel and irradiated for 20 min in microwave (Scheme 8) Cycloadduct 32 was obtained in a slightly lower yield (60%) than 29 (70%) but it was compensated by an almost quantitative removal of trimethylsilylethoxy group by TBAF To finish the total synthesis of rhododaurichromanic acids A and B, daurichromenic acid was irradiated with a low pressure mercury lamp for 5 d to afford a mixture of 1a (40%) and 1b (20%) This total synthesis could have a significant impact on the synthesis of natural chromenes/chromanes.
TM 5 og TMS 9 OOH SSa CaCl, HạO, NEts ) O R EtOH, microwave O SS ⁄
Wilson and co-workers have undertaken a slightly different approach toward synthesis of daurichromenic acid and its analogues (Scheme 9).29 Tietze’s conditions”’ were used to build the chromene core The difference is that diketo ester was used in place of diketone A mixture of diketo ester 33 and enal 28 was treated with 1,2- ethanediammonium diacetate at rt to give 34 in excellent yield Remarkably, cycloadduct
34 was the only regioisomer formed which could be attributed to the steric factors.Several other aldehydes and different diketo esters were used to give precursors for analogues of daurichromenic acid.
The most troublesome step in their synthesis was aromatization DDQ in refluxing benzene proved to be the best condition but still low yielding The use of a less reactive quinone, chloranil, failed to provide satisfactory results A stepwise dehydrogenation was also investigated Phenyl selenide derivative of 34 was treated with m-CPBA then 3,5- dimethoxyaniline to promote oxidation and elimination sequence”! but the yields were very low Aromatization of silyl enol ether derivatives of 34 with DDQ was also low yielding There will be more discussion later about this troublesome aromatization, as we were the first ones to use it to synthesize daurichromenic acid and rhododaurichromanic acids A and B.
Finally ethyl ester of daurichromenic acid 35 was successfully saponified using 5
M NaOH in DMSO upon heating at 80 °C for about 16 h This saponification protocol afforded significant quantities of daurichromenic acid and its analogues in good yields and avoided problematic decarboxylation reaction, as was seen in the previous synthesis by Jin.!$ Wilson’s methodology proved to be useful in preparing analogues of daurichromenic acid 36, 37, 38 which could potentially possess important biological activities (Figure 3).
Figure 3 Analogues of Daurichromenic Acid.
Previous Work: Hongoquercin A and B
Only one synthesis of hongoquercin A by Mori and co-workers was reported at the onset of our interest in this natural product.” Before Mori’s synthesis the absolute stereochemistry of hongoquercin A was not known which, in part, made it an interesting target for synthesis Oxidative degradation of the commercially available (-)-sclareol
(39), a method developed by Barrero,”” was used to make chiral enal 40 (Scheme 10).
The advantage of using this protocol is that the starting material is relatively inexpensive and it already has the desired stereochemistry.
OH OsOz/NalO¿, rOH OAc )© z5 93% -2) O3; then MeaS (93 %)
Oxidative degradation of (-)-sclareol started with exposing it to OsO,/NaIO, to give acetoxyaldehyde 41 Further degradation, to shorten the side chain of 41, was successfully accomplished via ozonolysis of its silyl enol derivative Corresponding acetoxyaldehyde was regioselectively eliminated by heating it with collidine to provide enal 40.
HO CH,Cla/MeOH (73%) SEMO
Aldehyde 40 was reduced to the corresponding alcohol which was, in turn, treated with phosphorus tribromide to provide bromide 42 (Scheme 11) Synthesis of the second coupling partner, phenolic acid part 44, started with iodination of commercially available ethyl orsellinate 43 Both of the hydroxyl groups were then protected as bis(2- trimethylsilylethyloxy)methoxy ether.
SEMO BuLi, CuCN, THF; 1) HạSOa, THF/EtOH (87%)
Coupling of 44 and 42 was successfully accomplished by converting 44 to its cuprate derivative and then treating it with 42 (Scheme 12) Removal of both SEM protective groups and exposure of the corresponding phenol to BF3-OEt2 in CH;C]¿ gave the desired tetracycle in 58% yield There was no spiral product with benzofuran ring system formed Ethyl ester of 8a was then hydrolyzed to finally give hongoquercin A.
Decarboxylation was not a problem in this case as it was seen in daurichromenic acid synthesis.’
SEMO 1) BuLi, CUCN, THF; 29 + 2) HạSO¿, THF/EIOH _
1) Hạ, Pd/C, EtOAc (93%) 2) KOH, THF/MeOH/H20 (96%)
3) AcaO, pyridine 4) KoCO3, MeOH/H20 (89%, 2 steps) AcO
Synthesis of hongoquercin B was achieved using the same approach (Scheme
13).' The lower part 46 is a known compound which was previously reported by Mori.”4
Coupling reaction between 46 and 44 and the removal of SEM protective groups was done in the same manner as before Upon treatment of compound 47 with BF:-OEt, it underwent the desired ring-closure to give 48, but isomer 49 was also formed.
Fortunately, these two regioisomers were readily separated by recrystallization from ethanol Finally, the synthesis of hongoquercin B was completed in 4 steps from compound 48 The absolute stereochemistry of hongoquercin B was established to be the same as hongoquercin A.
The new development of the methodology aimed at the enantioselective synthesis of chiral chromenes and chromanes needs to be discussed Several research groups synthesized chiral chromenes/chromanes or developed methodology towards enantioselective synthesis of chromene or chromane core’, as Scheme 14 demonstrates some representative examples.
OH NN OH (-)-DET, Ti(isopropoxide),, TBHP, F
HO ị 3 toi MeONa, MeOH, reflux 3h XXZ* “p-tol
Me Lọc | XC-Bu fe)
For example, Chandrasekhar uses a one-pot Sharpless asymmetric epoxidation, intramolecular epoxide opening with internal phenoxide anion to generate the chiral chromane 51.7* While Solladie employs chiral sulfoxide 52 in his asymmetric synthesis of Vitamin E.”*° Hoveyda employs Schrock’s Mo-catalyzed ring-closing metathesis to prepare optically pure 2-substituted chromane 55 in the first enantioselective total synthesis of the antihypertensive agent (S,R,R,R)-nebivolol ®° To discuss approaches towards chiral chromanoids comprehensively would be a great undertaken of its own and is beyond the scope of this thesis If some omitted researchers, and there are many, who were involved in such developments by some weird circumstances were to look into this dissertation, they would have to forgive an author for not discussing their contribution.
More recently, Trost and co-workers became interested in developing general methodology to construct chiral chromenes/chromanes.”° They used Pd-catalyzed intramolecular allylic alkylation of phenols to access chiral chromanes (Scheme 15).”° iS Pd, L* ức [OI Tứ ys
Trost’s methodology ought to be described in more detail because of their interest in daurichromenic acid and rhododaurichromanic acid A Scheme 16 illustrates synthesis of the intermediate 59 for the Pd-catalyzed intramolecular allylic alkylation Both phenolic groups of orcinol were protected as methyl ethers Compound 60 was obtained in a high yield after formylation and consecutive Wittig olefination of a bis-protected orcinol Hydroboration-oxidation sequence followed by iodination gave iodine 62 In order to couple alkyl iodide 62 and vinyl iodide Negishi coupling was used to give 63.
Monodemethylation of 63 was achieved using lithium propylthiolate in HMPA and allylic alcohol exposed to methyl chloroformate to provide allyl carbonate 64 which was found to be a more desirable leaving group for their feature Pd-catalyzed cyclization reaction.
OH 1) MezSO¿, KzCO3 (100%) OMe E270 OMe
2) nBuLi, TMEDA; the DMF (97%) ce then HO, NaOH OH
OH 3) PPhạ=CH; (98%) OMe 85% OMe
1) fBuLi, ZnCl;, -78 °C OMe OCO;Me OMe OH then TBSO ® OMe oc 1) nPrSLi (89%) oo `^ load -=————- —
OH 2) CICO2Me (95) OM Pd(dppf)Cls, nBuLi OMe @
Finally, they were ready for the key reaction to enantioselectively construct the chromane core (Scheme 17) To insure the best outcome of this reaction, the effects of additive, concentration, catalyst loading, ligands, leaving group, solvent, and temperature were systematically studied After much of the experimentation, it was found that without
AcOH the ee drops dramatically to 14% While temperature did not have any significant effect on the outcome of the reaction, however enantioselectivity and yield dropped slightly when reaction was run at 0 °C Chromane 65 was obtained in high yield and good
OMe OCOMe pd.dba;, CHạCl;, OMe
In order to proceed with the total syntheses of daurichromenic acid and rhododaurichromanic acid A, the side chain of 65 had to be elongated Various types of cross-coupling methods failed to lengthen the side chain An alternative strategy involved the addition of iodine 66 to chromene aldehyde 67 (Scheme 18) A more easily cleavable MOM protecting group was chosen instead of methyl ether Also MOM group could be an ortho-directing group for metalation in order to achieve the carboxylation the aromatic ring later in the synthesis The corresponding compound was hydroborated to give a 3:1 regioisomeric mixture of alcohols, the major being alcohol 68 The direct dehydrogenation of 68, by exposing it to DDQ and refluxing it in toluene to afford chromene core, failed Primary alcohol 68 was then protected with TBDPS The resulting silyl ether underwent a clean DDQ oxidation to give chromene 69 in 86% yield Silyl group was removed and the resulting alcohol was oxidized using TPAP to afford chromenic aldehyde 67.
OMe 1) C;H¿SH, NaH, DMF, 120 °C (94%) OMOM
O71 — 3) BHs-THF, THF, 0 °C to rt; oT on
65 then HạO;, NaOH, rt to 50 °C (94%)
1) TBDPSCI, DMAP (82%) 2) DDQ, PhH, reflux (86%)
> 1) TBAF, THF (90%) S oto < © ` 2) TPAP, NMO (98%) O OTBDPS
Finally, two building blocks 66 and 67 were coupled using Negishi conditions (Scheme 19) Vinyllithium derivative of 66 was added to aldehyde 67 which yielded allylic alcohol 70 as a mixture of two diastereomers, both of which were converted to carbonate 71 without separation The highly Pd-catalyzed regioselective reduction of both diastereomers of carbonate 71 provided the desired product 72 No scrambling or
Total Synthesis of Hongoquercin A se 80
Retrosynthetically, we envisioned two pathways leading to (+)-hongoquercin A
(Scheme 32) First approach would be to construct (+)-hongoquercin A using chiral enal
105 utilizing oxa-[3 + 3] cycloaddition with consecutive aromatization Our preliminary calculations indicated that the more stable cycloadduct 106 would have a desired stereochemistry at C8 In addition, having successfully made methyl] (+)-daurichromenic ester 86 we proposed the second route that would employ cationic polyene cyclization.
The latter approach will be discussed in the next Section.
$ \H 8a: (+}hongoquercin A 86: methyl daurichromenic ester formal
Synthesis of aldehyde 105, which is also a known natural product isolated from a
Japan”, turns out to be presedented.**”? Because enal 105 has been reported in the literature and we were hastily in need of starting material 105, to drive the effort toward total synthesis of hongoquercin A, the fastest rout was chosen The approach which involved cationic polyene cyclization was very attractive from that stand point It started with farnesol 88 which was protected as its acetate to give compound 107 To promote the olefinic cyclization cascade, 107 was treated with N,N-dimethylaniline tertiary amine complex of mercury(ID triflate (Scheme 33).** The intermediate mercury(II) triflate was then converted in situ to mercury(II) chloride followed by reduction with NaBH, to give diol 108 It was envisioned that diol 108 would be oxidized, followed by a removal of a hydroxyl group by refluxing the corresponding hydroxy aldehyde in toluene in the presence of catalytic amount of p-TsOH, giving enal 105.
OAc OH 1) Hg(OTf); : CaHaN(CHạ)s;
Unfortunately, the above route failed to provide enough diol 108 to even try oxidation and elimination sequence Besides it would lack stereospecificity Guided by these factors we were forced to explore the longer route toward the cycloaddition precursor, enal 105.
The second approach we have undertaken has been reported in the literature.”However, there was no experimental procedure in the original paper’ and we were also eager to improve the reported protocol by modifying the original procedure The advantage of the latter approach is that it is enantioselective, potentially higher yielding and more straightforward four-step synthesis The synthesis of the requisite oxa-[3 + 3] cycloaddition precursor, enal 105, is outlined in Scheme 34.
Commercially available (+)-sclareolide (109) was converted to a-hydroxylactone
111 by first treating it with KHMDS and then reacting the resulting anion with using
Davis oxaziridine 110.°° Compound 111 was formed as a single diastereomer although it had no further importance.
The use of potassium KHMDS instead of LDA greatly improved the yield of the oxygenating reaction It has been previously proposed that diisopropyl amine, the by- product of the deprotonation, competes with the enolate for the Davis’ sulfonyloxaziridine The use of Vedejs’ reagent, molybdenum peroxide-pyridine- hexamethylphosphoramide (MoOPH) has also been considered for installment of the a- hydroxyl group The disadvantage of this approach is the formation of the dimer as a side-product Indeed, when the MoOPH-mediated oxygenation was tried by Quideau et al on (-)-isosclareolide,*® the yield was lower (59% plus dimer in 24% yield) compared ọ 1) PhCH(OEt);, 160 °C Đo AS é ÀÈ ấm, Ph—S-N-G-Ph
The original procedure used a very expensive (+)-(10- camphorsulfonyl)oxaziridine to achieve this transformation To avoid the use of the expensive oxygenating reagent and also guided by the notion that the stereochemical outcome of this reaction is not essential, we decided to make 2-sulfonyloxaziridine 110 which is easily accessible from readily available starting materials.*’ Davis’ sulfonyloxaziridine 110 was made by condensing sulfonamide 112 with the protected benzaldehyde at 160 °C followed by epoxidation with m-CPBA in the presence of catalytic amount of BTEAC (Scheme 35).
To continue with the synthesis, a-hydroxylactone 111 was converted to triol 114 by reducing the lactone ring with LAH in refluxing THF (Scheme 36) A minor isomer, hydroxy lactol 113, was also formed in low yield Quideau also reported the formation of such by-product, moreover all of their attempts to further reduce it with either LAH or
DIBAL-H failed.*° However, Chackalamanni e/ al reports a quantitative conversion of the lactone moiety of 111 to triol 114 with LAH in refluxing THE.”?° Additionally, after quenching, a crude reaction mixture is dried with Na2SO, and filtered, first using simple gravity filtration followed by filtration thought he CeliteTM and aq work-up is avoided due to a great polarity of triol 93.
NalO, THF/HạO 99% p-TsOH, PhMe, reflux, 53%
Oxidative cleavage of the triol 114 with NalO¿ proceeded very smoothly and gave aldehyde 115 in 99% (Scheme 36) To finish synthesis of enal 105, tertiary alcohol was eliminated using p-TsOH and heat The yield of the above elimination could be improved by first converting alcohol 115 to the corresponding acetyl and then refluxing it in toluene At this point we were more interested in expediency and did not try to optimize elimination conditions In addition, enal 105 was found to be decomposing rapidly even at -10 °C under nitrogen atmosphere and so had to be used quickly.
With the precious intermediate 105 in hand, we were now ready to build the hongoquercin core using our formal oxa-[3 + 3] cycloaddition The stereochemistry of this key cycloaddition is ultimately controlled by the geometry of the substrate (Scheme37) The configuration of the stereogenic center adjacent to the oxygen depends on the thermal stability of cycloadduct 106 vs cycloadduct 106’ Based on simply PM3 was found to be 2.5 kcal/mol more stable thanl 06° Therefore 106 would form almost exclusively. ° 0.
: đãi : 6n electrocyclic lộ 6x electrocyclic ring-closure ring-closure
Attempts to generate iminium salt by heating 105 in the presence of piperidine and Ac;O followed by reacting with 5-methylcyclohexane-1,3-dione failed resulting in decomposition and recovery of some starting material Ultimately, after some experimentation, this key cycloaddition was accomplished by heating enal 105 and diketone 77 in the presence of piperidinium acetate in EtOAc using a sealed tube at 85 °C providing the desired cycloadduct 105 as a 1:1 mixture of two diastereomers at C19 in
LDA, THF, - 78 °C; then NCCO;Me 78% CO2Me
Compound 106 was then converted to its enolate derivative and added to the
Mander’s reagent furnishing -ketoester 116 as a complex mixture of isomers.
Stereochemistry of this acylation was irrelevant since it was followed by oxidative aromatization thus separation of diastereomers 116 was never attempted They gave separate peaks in LCMS Sometimes, especially if the reaction time was longer than usual and stoichiometry of methyl cyanoformate was not carefully controlled, isomer 117 was isolated In the previous synthesis of rhododaurichromanic acids A and B 1.5 equiv of Mander’s reagent was needed for the acylation reaction It was found that in the present case, stoichiometry had to be carefully kept at 1.1 equiv to avoid unwanted side reaction and to obtain compound 116 cleanly Fortunately, side-product 117 was cleanly converted to 116 with NaOH.
Aromatization of 116 with DDQ in benzene at 60 °C for 2 d provided 118 in 46% yield (Scheme 39) Treatment of silyl or methyl enol ether of 116 with DDQ or using higher temperatures gave only traces of the desired product Benzene was a much superior solvent than toluene, in this case Refluxing reaction mixture in toluene provided only trace amounts of product One of speculations is that toluene itself is more reactive towards DDQ oxidation than benzene, but in the case of previous synthesis of daurichromenic acid no such striking difference was observed Olefin at C9-C15 was also suspected to interfere with aromatization reaction making product 118 easier to decompose under reaction conditions Since it had to be eventually hydrogenated it was decided to reduce 116 before attempting DDQ aromatization This transformation was easily achieved by simple catalytic hydrogenation, stereochemistry of the corresponding hydrogenated product 119 was never determined, since it the starting material was a complex mixture of isomers However, compound 119, where alkene was absent, failed to aromatize under the above conditions.
CO,Me aq KOH, MeOH/THF;
Reduction of alkene 118 was best accomplished with TFA and TESH in refluxing
DCM which gave 120 as a single isomer in 99% yield (Scheme 40) Catalytic hydrogenation using Pd/C gave 120 as a 10:1 mixture of isomers The isomeric ratio, in both cases, was easily assigned by 'H NMR and refined by LCMS analysis Proposed mechanism of TFA/TESH reduction is outlined in Scheme 40 Protonation with TFA would give a mixture of 121 and 122, this process is possible reversible, which would equilibrate to the more stable intermediate 122 A hydride transfer would have to be a rate limiting step, and only capture a predominant species 122, since only a single isomer is formed Also if hydride attack happens at a tertiary cation reaction stops and no desired down Long reaction time and a need for elevated temperature support the steric and a slow hydride transfer arguments The stereochemistry of reformed center adjacent to oxygen is controlled by the substrate as in the case of [3 + 3] cycloaddition.
Finally, methyl ester 120 was saponified using aqueous KOH in refluxing
THF/MeOH for 2 h to afford (+)-hongoquercin A (Scheme 40) Decarboxylation was seen if the reaction was allowed to proceed for more than three hours To compare synthesized 8a with the reported natural product, it was converted to its mono potassium salt which gave spectroscopic data identical to reported values © It was reported by Mori et al that IR, 'H and '°C NMR spectra were easily affected by the pH of the solution used for analysis The salt of 8a gave different spectroscopic data to that of free 8a, complicating identification After analyzing authentic sample of natural 8a they were able to prove that their synthetically hongoquercin A is identical to the natural one. Furthermore, spectroscopic data in the original isolation paper appears to be of potassium salt of hongoquercin A To be absolutely sure we compared both free 8a and its potassium salt to the reported values.
Polyene Cyclization and Cationic [2 + 2] Cycloaddition
The idea that naturally occurring compounds with multiple rings could be biosynthetically made from linear compounds having double bonds positioned in the way that successive ring-closure could happen is known as Stork-Eschenmoser hypothesis and was proposed in 1955.°` In tums out, that this type of cascade cyclization is also possible in the laboratory and could be accomplished without enzymes (Scheme 41) It is often called polyene cyclization or olefinic cyclization.
Over the years numerous reagents were used to promote polyene cyclization.Natural occurring terpenoids and steroids have been the most attractive target for polyene cyclization The reaction that constructs several bonds are at once stereoselectively possess a very important synthetic method Using epoxide as an anchor for this type of cyclizations has been also widely adopted (Scheme 41).
The construction of (+)-hongoquercin A (8a) could be envisioned via such cationic polyene cyclization of 86 (Scheme 42) The stereochemical outcome of this cyclization is dictated by the configuration of methyl daurichromenic ester (86), enantiomer of the natural configuration is shown, according to the Stork-Eschenmoser hypothesis Using this hypothesis we predict that the stereochemistry of the cyclization product will be the same as the stereochemistry of (+)-hongoquercin A (8a). Alternatively, epoxide 126 could be cyclized giving rise to the hongoquercin B skeleton (127) Hongoquercin A could be obtained via Barton’s deoxygenation protocol followed by hydrogenation and saponification The issue of the uncertainty regarding stereochemical assignment of daurichromenic acid has been brought in the introduction part of this Chapter It would be much more opportunistic for us if the stereochemistry of the natural product would be revised from what it was originally reported This way the revised configuration of the natural product would be used for the polyene cyclization
The proposed cationic polyene cyclization is unique given the position of the third alkene in the chromene core, so we named it exo-polyene for the third olefin is three carbons away from the second olefin (Scheme 43) Despite vast precedents in the endo- type polyene cyclization including those reported by Omura, Smith, and Parker using trienes 128,” it is uncertain of its outcome and more importantly, the regioselectivity in the second bond formation, C9-C10 versus C15-C10.
128 exo-type polyene endo-type polyene : cyclization cyclization Smith/Omura/Parker
Initially we attempted to use various Lewis acids or mercury reagents to cyclize
86 or its precursors 89 or 93 (Scheme 44) Having successfully synthesis hongoquercin A it would be easy to intercept one of its intermediates even if it forms in a very low yield. The plan was to screen several reaction conditions and look for desired cyclized products in the crude 'H NMR.
Lewis Acids or Hg?* no cyclization
Lewis Acids or Hg2* no cyclization
Unfortunately all our attempts to our great and growing frustration failed. Decomposition was seen in some cases and no reaction in others We screened several Lewis acids including: BF3-EtO2, AlCls, InCh, ZnBr2, CuCh, FeCh, SnClz, MeAICl, TICH:(O-¡Pr) Different solvents were also tried including a polar CH3NO2 which in some cases was shown to greatly affect the outcome of polyene cyclization by stabilizing a positive charge developed during the reaction Experimenting with different temperature, concentration or stoichiometry did not have any positive effect.
As an alterative, Hg(OAc); and Hg(OTf)2 were tried Again no cyclization was seen However, when ketone 89 was exposed to Hg(OTf), an unidentifiable product was formed When it was treated with NaCl, to convert mercury triflate group which was still on the molecule to mercury chloride for the ease of chromatography, it gave a product which had mercury chloride on it, as was seen by LCMS It was cleanly reduced with
NaBH, but we still could not identify the product It had a mass higher than the desired product 106 by 18, but the rest of the spectroscopic data did not agree to any of our proposed structures This unidentifiable product was obtained as oil and all our attempts to crystallize it were to no avail.
Finally, we tried TFA At first there was no cyclization seen when 1-10 equiv of TFA was used but when 86 was exposed to 30 equiv of TFA some desired product 118 was seen in NMR along with lots of unreacted starting material Scheme 45 shows some of the conditions that we tried Along with 118 another product 129, unknown at that time, was isolated in a greater yield.
118 129 hongoquercin, A rhododauriochromenic acid A derivative Scheme 45
Table 1 summarizes reaction conditions that were explored Cyclization was usually complete after 1 h Addition of TFA was usually done at -78 °C and the reaction was allowed to stir for 30 min at which point dry-ice/acetone bath was removed and reaction was allowed to warm up to rt for additional 30 min Decomposition of 129 was seen when reaction time was longer than 3 h and also when reaction was performed at higher temperatures Molecular sieves were also added to the reaction However, when DCM was not dry before the reaction it did not seem to have any effect on the outcome. initiator (equv) solvent temp time Yields (97 : 106) TFA (5.0-10.0) CHUCI, -78 to rt 0.5-48 trace trace
TFA 1:3 CH,CI, -78 to rt 1 19 35
TFA 1:5 CH,Cl, -78 to rt 3.5 16 36
TFA 1:5 CH,Cl, -78 to 0 17 23 trace TFA/TFAA 1: 20 CH,CI, O tort 0.5 19 35
TFA 1: 20 CH,Cl, 0 to rt 3 17 27
Table 1 Reaction conditions for polyene cyclization.
Few words have to be said about the crude 'H NMR of this reaction The upfield region of crude NMR was not useful at all as it was almost not comprehensive Two of cyclobutane protons of 129 were clearly resolved but olefinic and aromatic region was very crowded and few information could be deduced from it The most useful were phenolic peaks around 12 ppm, they were very clearly resolved This was used to track reaction progress and to identify ratios of 118 and 129 The above information was also used to detect several by-products forming in very small amounts As much as four by- products were seen Unfortunately separation of 118 and 129 proved to be very difficult and screening of chromatographic fractions did not produce any other relevant products or related stereoisomers, and thus, the formation of 118 and 129 was also consistently highly diastereoselective.
Interconversion between 118 and 129 was not observed when individual pure samples were subjected to the reaction conditions besides again a more facile decomposition of 129 was seen While 118 was assigned with confidence, spectroscopically it exactly matched previously made (+)-118, we had no idea what 129 was except the ratio of 118:129 was consistently ~ 1:2 in all cases.
Initial attempts to crystallize 129 were not successful It was either amorphous, formed foam or crystallized into very thin needles which were not suitable for X-ray analysis More efforts were needed to assign 129, and ultimately, X-ray structure of the para-bromobenzoate derivative 130 was obtained (Figure 4).
We were very surprised to find that it possesses a caged tricyclic structural motif with a cyclobutane, which closely matched that of rhododaurichromanic acid A A closer reexamination the 'H NMR of 129 also confirmed that it had all the relevant resonances except the lone olefin resonance Moreover, it indicated that 129 was forming as a single isomer and rhododaurichromanic acid B derivative was not forming. ® =) tol, 100 °C
To our big surprise elimination of triflouroacetic group was not trivial In many reported cases elimination of this sensitive group is seen after exposing compound to silica gel but in our case it survived extensive chromatography Bases, such as NaOH, EtN, DBU, K2CO; or stronger once, such as LDA and NaH failed to accomplish the desired transformation Eventually, a two-step protocol was used to convert triflouroacetate to olefin (Scheme 46) Compound 129 was allowed to react with KạCOa at rt for 16 h in MeOH Tertiary alcohol 131 was obtained almost quantitatively. Spectroscopically the formation of 131 was hard to detect The only difference between
Approach Toward Hongoquercin B
The success of applying polyene cyclization to the synthesis of hongoquercin A led us to explore the possibility of using this cyclization in the synthesis of hongoquercin
B It was mentioned in the introduction, that attempts have been made to use polyene cyclization to construct hongoquercin B core!? from the corresponding linear precursor possessing epoxide functionality but no further inside has been given As in the case of hongoquercin A, this is a unique exo-type cyclization, while endo-type cyclizations are precedent” (Scheme 52).
Ox LOW Ar endo “SS | SJ} OH
OH CHCl, -10 °C CO¿Me or mixture of epoxides or bis-epoxidation
1) NBS, THF/H;O, 0 °C in both cases
Attempts to epoxidize 86 or its precursors 89 and 93 produced either a mixture of epoxides or bis-epoxidized products Therefore, a different approach, where epoxide moiety had to be introduced at the early stages, was chosen Epoxide 148 was prepared from commercially available farnesol in 4 steps by following a reported procedure
(Scheme 53).' Farnesol was first protected as its acetate derivative using the standard conditions, giving the corresponding acetate in almost quantitative yield The bromohydrin 149 was preferentially formed at the terminal olefin by reacting farnesol acetate with NBS Finally formation of the epoxide was accomplished by treating the bromohydrin 149 with methanolic K,CO3 at room temperature which also served to remove the acetate group Oxidation of the resulting epoxyfarnesol was first attempted with SO3-pyridine complex but it did not give desirable yields Alternatively, MnO was used to convert epoxyfarnesol into epoxy enal 148.
OH = 1) Ac, 0, pyridine, 99% OAc tt, 99%
A formal oxa-[3 + 3] cycloaddition between epoxy enal 148 and 5-methyl-1,3- cyclohexanedione (77) proceeded to give cycloadduct 150 in 31% yield (Scheme 54). Higher temperatures resulted in only trace amounts of the desired cycloadduct 150 The yield was improved with piperidinium acetate to 55% and use of Lewis acids did not produce any product Compound 150 was then converted to B-keto ester 151 with Mander’s reagent. piperidinium acetate, Na,SO, a LDA, MeCQ2CN
To our biggest disappointment every aromatization conditions we tried led to no product In some cases, some of the aromatized but half-cyclized product 152 was seen(Scheme 55) In addition, after extensive experimentation, we could not achieve the formation of the tetracyclic skeleton of hongoquercin B Several Lewis acids were tried:FeCl;, AIC, TiCl4, Cp2ZrCl,, AlMe3, EtzAICl, SnCly Other reagents such as TFA, NT and HOAc were also tried In most cases, the only product we were isolating, in quite good yield, was a half-cyclized compound 153 or 154 containing a bridged tetrahydrofurane ring system (Scheme 55) Longer exposure of these bridged compounds to Lewis acids, in hopes of completing the formation of a tetracycle, led to slow decomposition of the starting materials The newly formed bridge in 152, 153 or 154 is racemic, since the starting epoxide is racemic These results confirmed our fears about unexplored exo-type polyene cyclization, since formation of tetrahydrofurane ring was far more favorable than the formation of the second six-membered ring.
Undiscouraged by the above results, we decided to pursue the second route which would involve synthesis of cycloaddition precursor 155 via the polyene cyclization of epoxy farnesal 148, unfortunately this did not work either (Scheme 56).
We also briefly considered a more straightforward synthesis of 155 This idea was rejected; since synthesis of 155 would be too long and time-consuming for this synthesis to be attention-grabbing or innovative Besides, we did not feel that synthesis of hongoquercin B would be synthetically important as it possesses very weak biological activity and its synthesis would follow along the same path of hongoquercin A After some deliberation we abandoned synthetic efforts towards hongoquercin B all together.
Overall Conclusion ch ha 107
The describe efforts represent the first applications of our formal [3 + 3] cycloaddition approach in the synthesis of natural chromenes and chromanes They were illustrated with a concise total syntheses of (+)-rhododaurichromanic acids A and B, methyl ester of (+)-daurichromenic acid and (+)-hongoquercin A In addition, an unusual polyene cyclization and a unique cationic cyclobutane formation was described which could explain the biosynthetic relationship between rhododaurichromanic acids A and daurichromenic acid It also represented a divergent path towards hongoquercin A and rhododaurichromanic acids A We are very optimistic about widening the scope of cationic [2 + 2] cycloaddition There are rigorous synthetic efforts underway in our laboratory to explore this unusual reaction More effort has to be devoted to develop enantioselective approach towards chromene/chromane synthesis Such methodology could be utilized to make a chiral precursor for the cationic [2 + 2] cycloaddition reaction, to shed light on the exact mechanistic details of this unusual process.
There are definitely a lot of natural products that contain chromene or chromane core and more are being discovered Synthesis of such compounds or its derivatives might be interesting from the stand point of the scientist who is attracted by their biological activity We hope that our synthetic work in the area of chromanoid chemistry would prove useful to such individual.
References HS HH teen enna ene Ki kế 108
Muhammad, I.; Li, X-C.; Dunbar, D C.; El Sohly, M A.; Khan, I A J Nat Prod 2001, 64, 1322-1325.
Vitamin E; Machlin, L J., Ed.; Marcel Dekker: New York, 1980.
(a) Seeram, N P.; Jacobs, H.; McLean,; Reynolds, W.F Phytochemistry 1998, 49, 1389-1391 (b) Tanaka, T.; Asat, F.; Inuma, M Phytochemistry 1998, 49, 229-232.
(a) Hirai, K.; Suzuki, K T.; Nozoe, S Tetrahedron, 1971, 27, 6057 (b) Nozoe, S.; Suzuki, K.Y. Tetrahedron Lett 1969, 29, 2457.
(a) Roll, D M.; Manning, J K.; Carter, G T, J Antibiot 1998, 57, 635-639 (b) For fermentation studies see: Abbanat, D A.; Singh, M P; Greenstein, M J Antibiot 1998, 51, 708-714.
Kashiwada, Y.; Yamazaki, K.; Ikeshiro, Y.; Yamagishi, T.; Fujioka, T.; Mihashi, K.; Mizuki, K.; Cosentino, L M.; Fowke, K.; Morris-Natschke, S L.; Lee, K.-H Tetrahedron 2001, 57, 1559- 1563.
Chinese Material Medica; Jiangsu New Medical College, Ed.; Shanghai People’s Pub House: Shanghai, 1997, p 2506.
Jpn Kokai Tokyo Koho, JP 82-28080, 1982.
(a) Cardillo, G.; Cricchio, R.; Merlini, F.; Nasini, G Gazz Chem Ital 1969, 99, 308-315 (b) Hellwig, V.; Nopper, R.; Mauler, F.; Freitag, J.; Liu, J.; Ding, Z.; Stadler, M Arch Pharm 2003,
(a) Ishil, N.; Takahashi, A.; Kussano, G.; Nozoe, S Chem Pharm Bull 1988, 36, 2918-2924 (b) Mahiou, V.; Roblot, F.; Hocquemiller, R.; Cave A J Nat Prod 1995, 58, 324-328.
(a) Iwata, N.; Wang, N.; Yao, X.; Kitanaka, S J Nat Prod 2004, 67, 1106-1109 (b) Goto, T.; Kakisawa, H.; Hirata, Y Tetrahedron 1963, 19, 2079 (c) Muckensturm, B.; Diyani, F.; Reduron, J.-P.; Hildenbrand, M Phytochemistry 1997, 45, 549.
Tsujimori, H.; Mori, K Biosci Biotechnol Biochem 2000, 64, 1410-1415.
For a more detailed review and applications of the formal [3 + 3] cycloaddition in natural product synthesis see: Hsung, R P.; Kurdyumov, A V.; Sydorenko, N Eur J Org Chem 2005, 23-44. Kurdyumov, A V.; Hsung, R P.; Ihlen, K.; Wang, J Org Lett 2003, 5, 3935-3938.
Y Kang, Y Mei, Y Du, Z Jin, Org Lett 2003, 5, 4481-4484.
Saimoto, H.; Yoshida, K.; Murakami, T.; Morimoto, M.; Sashiwa, H.; Shigemasa, Y J Org.Chem 1996, 61, 6768.
Roush, W, R.; Coffey, D S.; Madar, D J J Am Chem Soc 1997, 119, 11331.
Hu, H.; Harrison, T J.; Wilson, P D J Org Chem 2004, 69, 3782-3786.
Tsujimori, H.; Bando, M.; Mori, K Eur J Org Chem 2000, 297-302.
Barrero, A F.; Alvarez-Manzaneda, E J.; Chahboun, R Tetrahedron Lett 1997, 38, 8101-8104. Mori, K.; Koga, Y Liebigs Ann Chem 1995, 1755-1763.
(a) Chandrasekhar, S.; Reddy, M V Tetrahedron, 2000, 56, 6339-6344 (b) Hubscher, v J.; Barner, R Helv Chim Acta 1990, 73, 1068-1086 (c) Solladie, G.; Moine, G J Am Chem Soc.
Am Chem Soc 1998, 120, 8340-8347 (e) Uozumi, Y.; Kato, K.; Hayashi, T J Am Chem Soc.
Trost, B M.; Shen, H C.; Dong, Li.; Surivet, J.-P.; Sylvain C J Am Chem Soc 2004, 126, 11966-11983.
(a) A de Groot, B J M Jansen, Tetrahedron Lett 1975, 16, 3407-3410 (b) P F Schuda, W A. Price, J Org Chem 1987, 52, 1972_ 1979.
Kurdyumov, A V.; Hsung, R P.; Ihlen, K.; Wang, J Org Lett 2003, 5, 3935-3938
Mancuso, A J.; Huang, S.-L.; Swern, D J Org Chem 1978, 43, 2480. a) Becker, R S.; Michl, J J Am Chem Soc 1966, 88, 5931 b) Parker, K A.; Mindt, T L Org. Lett 2001, 3, 3875 c) Ahmed, S A.; Tanaka, M.; Ando, H.; Iwamoto, H.; Kimura, K Zur J Org. Chem 2003, 2437.
Toyota, M.; Oorso, Y.; Kusuyama, T.; Asakawa, Y Phytochemistry 1994, 35, 1263.
(a) Hua, D H.; Huang, X.; Chen, Y.; Battina, S K.; Tamura, M.; Noh, S K.; Koo, S L; Namatame, I.; Tomoda, H.; Perchellet, E M.; Perchellet, J.-P J Org Chem 2004, 69, 6065~
6078 (b) Chackalamannil, S.; Wang, Y.; Xia, Y.; Czarniecki, M Tetrahedron Lett 1995, 36, 5315-5318.
Nishizawa, M.; Takenaka, H.; Hayashi, Y J Org Chem 1986, 51, 806-813.
Davis, F A.; Vishwakarma, L C.; Finn, J J Org Chem 1984, 49, 3243-3244.
Quideau, S.; Lebon, M.; Lamidey, A-M Org Lett 2003, 4, 3975-3978.
(a) Davis, F A.; Lamendola, J., Jr.; Nadir, U.; Kluger, E W.; Sedergran, T C.; Panuto, T W.;Billmers, R.; Jenkins, R., Jr.; Turchi I J.; Watson, W H.; Chen, J S.; Kimura, M J Am Chem.Soc 1980, 702, 2000-2005 (b) Davis, F A.; Stringer, O D J Org Chem 1982, 47, 1774-1775.(a) Stork, G.;Burgstahler, A W J Am Chem Soc 1955, 77, 5068 (b} Eschenmoser, A.; Ruzicka,L.; Jeger, O.; Arigoni, D Helv Chim Acta 1955, 38, 1890
(a) Handa, M.; Sunzazuka, T.; Sugawara, A.; Harigaya, Y.; Otoguro, K.; Omura, S J Antibiot.
2003, 56, 730 (b) Smith, A B III; Kinsho, T.; Sunazuka, T.; Omura, S Tetrahedron Lett 1996,
37, 6461 (c) Tomoda, H.; Tabata, N.; Nakata, Y.; Nishida, H.; Kaneko, T.; Obata, R.; Sunzazuka, T.; Omura, S J Org Chem 1996, 61, 882 (d) Parker, K A.; Resnick, L J Org Chem 1995, 60,
(a) Gassman, P G.; Chavan, S P.; Fertel, L B Tetrahedron Lett 1990, 31, 6489 (b) Gassman, P. G.; Lottes, A C Tetrahedron Lett 1992, 33, 157.
Corey, E J.; Kania, R 8 J Am Chem Soc 1996, 118, 1229-1230.
Experimental Procedures c se 111
General Methods: Column chromatography was performed on Bodman silica gel (60 A, 230-400 mesh) THF was distilled over sodium/benzophenone under nitrogen.
Dichloromethane, toluene, acetonitrile, diisopropyl amine, triethylamine, and benzene were distilled from calcium hydride under nitrogen Methanol was distilled from magnesium and stored under nitrogen over 3 A molecular sieves N,N- dimethylformamide was distilled under vacuum from calcium hydride and stored over 4
A molecular sieves Flasks were flame dried under vacuum and purged with nitrogen before use TLC plates were silica (Whatman, polyester backed) and were visualized with UV (254 nm) and either anisaldehyde or permanganate stains IR spectra were recorded on NaCl plates using a Midac M2000 FTIR IR intensities are reported as follows: w = weak, m = medium, s = strong, vs = very strong, br = broad 500 MHz 'H spectra were recorded on a Varian Inova spectrometer; 300 MHz spectra were recorded on a Varian Unity or Varian Inova instruments and are referenced to TMS at 6 0.00 ppm. Other NMR solvents are referenced to residual solvent NMR peak multiplicities are reported as follows: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br broad '°C spectra were recorded on a Varian Inova spectrometers at 125 and 75 MHz and are referenced to the center chloroform peak at 6 77.23 ppm Electrospray mass spectra were recorded on a Bruker Biotof If ESI-TOF/MS using either PPG or PEG standards as high resolution calibrants Unless noted, all reagents (Acros, TCI, Aldrich) were used as received Crystallographic analysis was performed at the University of Minnesota X-Ray lab.
Compound 87 Farnesol (4.39 g, 0.0197 mmol) was dissolved in 80 mL of CH¿C]; The solution was cooled to 0°C, and DMSO (15.47 ml, 0.217 mol) and EtạN (13.73 mL,
0.988 mol) were added in this order to the reaction mixture SO3-pyridine (12.57 g, 0.079 mol) was added slowly over 10 min After 50 min reaction was worked-up with 3% aq
HCl, sat aq NaCl, and dried with NaSO, The crude product was purified using flash silica gel column chromatography (gradient eluent: 0-10% EtOAc in hexane) to give 4.09 g of farnesal 87 (94%).
Ry = 0.43 (20% EtOAc in Hexane); clear oil;
(m, 2H), 5.88 (dd, 1H, J =1.2, 8.1 Hz), 10.00 (d, 1H, J = 8.4 Hz); C NMR (75MHz,
164.0, 191.3; IR (neat) cm! 2853s, 2767m, 1677s, 1632s, 1611s, 1443s, 1381s; mass spectrum (APCI): m/e (% relative intensity) 221 (50) (M + H)”, 204 (20), 203 (100), 147
Method I Farnesal 87 (3.77 g, 0.0171mol) was added to the solution of piperidine (3.70 mL, 0.0374 mol) in EtOAc (30 mL) The solution was cooled to 0°C and stirred for 5 min After which, acetic anhydride (3.7 mL, 0.0392 mol) was added dropwise The flask was transferred into the 90°C oil bath and stirred for 1 h The solution was then cooled down to room temperature and cannulated to a sealed tube that contained 5-methyl- cyclohexane-1,3-dione (3.23 g, 0.0256 mol) in toluene (70 mL) It was flushed with N; and heated at 90°C for 18 h The crude product was concentrated and purified using flash silica gel column chromatography (gradient eluent: 0-10% EtOAc in hexane) to give 3.91 g of 89 (70%).
Method II A mixture of 87 (5 g, 0.022 mol), 5-methyl-cyclohexane-1,3-dione (77) (4.28 g, 0.0339 mol), piperidinium acetate (0.022 mol) and Na2SO, in dry ethyl acetate (100 mL) was stirred for 16 hours at room temperature Reaction mixture was filter through the Celite The crude product was concentrated and purified using flash silica gel column chromatography (gradient eluent: 0-10% EtOAc in hexane) to give 6.13 g of 89 (85%). Method III To the ice-cold solution of 87 (1 g, 0.0044 mol), 5-methyl-cyclohexane-1,3- dione (77) (0.8 g, 0.00678 mol), in dry CH;C]; (25 mL) were added 4 A MS, followed by the addition of Zn(OTf)2 (0.7942 g, 0.0022 mol) The reaction mixture was stirred for 5 d at 0°C at which point it was poured into water The reaction was worked-up with sat aq NaHCO; until basic, washed with sat aq NaCl and dried over Na;SOa The crude product was concentrated and purified using flash silica gel column chromatography (gradient eluent: 0-10% EtOAc in hexane) to give 0.95 g of 89 (66%).
Ry = 0.29 (20% EtOAc in Hexane); yellow oil;
'H NMR (300MHz, CDCl;) Two isomers 5 0.98 (s, 3H), 1.00 (s, 3H), 1.26 (s, 3H), 1.31
J = 10.2 Hz), 5.11 (d, 1H, J = 10.2 Hz), 6.35 (d, 2H, J = 10.2 Hz); !'C NMR (75MHz, CDCI;) Two isomers 5 16.0, 17.7, 21.0, 21.1, 22.2, 22.6, 25.7, 26.6, 27.3, 27.6, 28.2,
Compound 93 n-BuLi (0.0131 mol) was added dropwise to a solution of diisopropylamine (1.17 ml, 0.0160 mol) in THF (15 mL) at -78°C The mixture was warmed to 0°C and stirred for 1 h It was then cooled down to -78°C and compound 89(3.90 mg, 0.0119 mol) was added dropwise The reaction was stirred at -78°C for | h, and
— after which, it was transferred to a solution of methyl cyanoformate (1.70 mL, 0.0237 mol) in THF (15 mL) via cannula After stirring for 10 min at -78°C and 10 min at 0°C the reaction was quenched with ice-cold water The reaction mixture was worked-up with 3% HCl, sat aq NaCl, and dried over NaSOa The crude product was concentrated and purified using flash silica gel column chromatography (gradient eluent: 0-5% EtOAc in hexane) to give 3.27 g of 93 (71%).
Ry = 0.35 (20% EtOAc in Hexane); pale-yellow oil;
'H NMR (300MHz, CDC];) Mixture of major isomers 6 1.06 (s, 3H), 1.08 (s, 3H), 1.34
(d, 2H, J = 10.2); !'C NMR (75MHz, CDCl;) Mixture of major isomers & 16.0, 17.7,
1650s, 1596s, 1416s; mass spectrum (APCI): m/e (% relative intensity) 387 (100)
OH we Z CO2Me fe)
DDQ Method A solution of 93 (0.89 g, 2.282 mmol) and DDQ (0.63g, 2.760 mmol) in toluene (50 mL) was refluxed overnight The reaction mixture was filtered through the celite followed by a standard aqueous work-up The crude product was concentrated and purified using flash silica gel column chromatography (gradient eluent: 0-10% EtOAc in hexane) to give 0.39 g of 86 (44%).
Via Selenoxide A solution of 93 (100 mg, 0.260 mmol) in dry THF (1 mL) was added dropwise to the freshly prepared solution of diisopropylamine (0.712 mmol) in THF (1 mL) at -78°C The mixture was stirred for 15 min after which a solution of phenylseleny] chloride (135 mg, 0.712 mmol) in THF (0.5 mL) was added fast The mixture was allowed to warm to -40 °C over 15 min Excess base was neutralized with benzoic acid (36.5 mg, 0.3 mmol) Then solution of m-chlorobenzoperoxic acid (160 mg, 0.821 mmol) in dichloromethane (3 mL) was added The reaction was allowed to stir for additional 1 h and the temperature was kept below -38 °C 3,5-dimethoxyaniline (73.4 mg, 0.498 mmol) in THF (1 mL) was added The reaction mixture was allowed to warm up to rt and stirred fro additional 30 min The reaction mixture was diluted with MTBE (5 mL) worked-up with 3% HCl, sat aq NaHCO¿, sat aq NaCl, and dried over NaSOa The crude product was concentrated and purified using flash silica gel column chromatography (gradient
Via Silyl Enol Ether To a solution of 93 (50 mg, 0.130 mmol) in dry CH¿C]; (1 mL) was added dropwise HMDS (0.032 mL, 0.149 mmol) at rt The reaction mixture was allowed to stir for 2 h under nitrogen at rt The solvent was evaporated in vacuo and a solution of DDQ (61 mg, 0.272 mmol) in toluene (5 mL) was added via cannula The mixture was refluxed for 5 h The crude reaction mixture was worked up analogous to
DDQ Method described above to give trace of 86.
Via Bromide To a solution of 93 (100 mg, 0.260 mmol) in dry THF (2 mL) at 0 °C was added NBS (47 mg, 0.260 mmol) as one portion The mixture was allowed to warm to rt and stirred overnight Thee solvent was evaporated in vacuo and DMF (2 mL), HạO (2 mL) and a “tip-of-spatula” of LiCl was added The mixture was refluxed for 3 h Crude
NMR showed trace amount of 93 and the extensive decomposition.
Ry = 0.59 (20% EtOAc in Hexane); clear oil;
158.0, 159.9, 172.5; IR (neat) cm’! 2968s, 2924s, 2856m, 1653s, 1620s, 1566s, 1441s, 1416m; mass spectrum (APCI): m/e (% relative intensity) 385 (100) (M+H)*, 354 (20),
Confluentin (13) To a solution of 86 (10 mg) in THF (1 mL) and MeOH (0.5 mL) was added LiOH (1 mg) in H2O (1 mL) The reaction was allowed to stir for 16 h at 60 °C. The reaction was acidified with 2% HCl, followed by extraction with CH2Cl The combined organic fractions were washed with sat aq NaCl, dried over NaSO, and concentrated Flash silica gel chromatography (10% ethyl acetate in hexane) gave 2.5 mg of 13 (23 %) Spectroscopically identical to to the one reported: (a) Cardillo, G.; Cricchio, R.; Merlini, F.; Nasini, G Gazz Chem Ital 1969, 99, 308-315.
Compound 99a/b Methyl (+)-daurichromenic ester 86 (56.6 mg, 0.147 mmol) was transferred into a 100-mL pyrex round bottom flask and dissolved in hexanes (30 mL).The reaction mixture was irradiated with medium-pressure mercury lamp for 65 h at room temperature while the lamp was placed 5 cm away from the flask The UV lamp was positioned in a standard photochemical emersion well and the well was cooled with chilled water The crude product was purified using flash silica gel column chromatography (gradient eluent: 0-2% EtOAc in hexane) to give 44.6 mg of 99a/b(79%).
Ry = 0.65 (20% EtOAc in Hexane); white powder;
(75MHz, CDC];) 5 14.6, 17.4, 17.5, 22.6, 23.3, 24.0, 25.5, 25.6, 25:6, 26.0, 26.6, 27.3, 29.9, 30.6, 35.4, 35.7, 37.8, 38.4, 38.8, 40.2, 42.0, 42.2, 44.3, 46.1, 46.1, 51.6, 84.1, 84.2, 104.3, 104.4, 109.3, 109.4, 113.0, 125.0, 125.1, 130.4, 130.8, 139.9, 158.0, 158.1, 162.9, 163.1, 172.4; IR (film) cm! 2952s, 2864s, 1648s, 1621s, 1576s, 1452s, 1408s; mass spectrum (APCI): m/e (% relative intensity) 385 (70) (M+H)”, 354 (25), 353 (100), 305
Compound 100 A solution of 86 (126 mg, 0.326 mmol) in hexanes (350 mL) was irradiated by medium-pressure mercury lamp using a standard photochemical reaction vessel and using Vicor filter, for 6 h at -78°C The crude product was purified using flash silica gel column chromatography (gradient eluent: 0-2% EtOAc in hexane) to give 15.6 g of 99a/b (12%) and 7.2 g of 100 (5.7%).
Ry = 0.15 (2% EtOAc in Hexane); pale-yellow oil;
131.6, 139.7, 147.7, 157.8, 162.2, 172.5; IR (film) cm! 2962s, 2928s, 2362ws, 2341ws,
1648s, 1653s, 1585s, 1445s; mass spectrum (APCI): m/e (% relative intensity) 385 (100)
Mixture of (+)-rhododaurichromanic acid A and (+)-rhododaurichromanic avid B (1a/b) To a solution of 99a/b (40 mg, 0.104 mmol) in methanol (6 mL) and THF (6 mL) was added 6 M aq NaOH (3 ml) After stirring at room temperature for 48 h, the reaction was acidified with 2% aq acetic acid, followed by extraction with equal volumes of CH2Cl, The combined organic fractions were dried washed with sat aq NaCl, dried over NaSO¿ and concentrated Flash silica gel chromatography (gradient eluent: 0-20% EtOAc in hexane) gave 27 mg of 1a/b (94%).
Ry = 0.35 (20% EtOAc in Hexane); white powder;
131.2, 141.9, 159.5, 164.2, 164.5, 176.6; IR (film) cm” 3406brs, 2962w, 2930w, 1621m, 1572w, 1455ms, 1366w; mass spectrum (APCI): m/e (% relative intensity) 371 (M+H)”
HPLC separation of l1a/b Compounds la/b (16.2 mg) were separated by semi- preparative HPLC (pump: Agilent Technologies 1100 Series; column: Alltech
Econosphere C18 10 p 250x22; elution with [1% aqueous acetic acid] : CH3CN with a ratio of 20 : 80 CH3;CN was removed in in vacuo and product was extracted from water using CH;CL¿ Evaporation of solvent afforded 5.5 mg of 7a (34%) and 4.5 mg of 7b(28%).
Ry = 0.36 (20% EtOAc in Hexane); white powder; mp 168-170 °C;
'H NMR (500MHz, CDC]:) 6 0.77 (s, 3H), 1.40 (s, 3H), 1.65 (s and m, 5H), 1.71 (s and m, 4H), 1.86 (ddd, 1H, J = 4.5, 12, 12 Hz), 1.90-1.96 (m, 3H), 2.06-2.13 (m, 1H), 2.50 (m, 1H), 2.53 (s, 3H), 2.58 (dd, 1H, J = 8.0, 10.0 Hz), 3.11 (d, 1H, J = 9.5 Hz), 5.18 (t,
125.2, 130.2, 141.8, 159.4, 164.5, 176.3; IR (film) em! 3433brs, 2967w, 2929w, 1621m, 1572w, 1460ms, 1366w; mass spectrum (APCI): m/e (% relative intensity) 371 (M+H)”
Ry = 0.34 (20% EtOAc in Hexane); white powder; mp 159-161 °C;
141.8, 159.5, 164.2, 176.2; IR (film) cm’ 3435brs, 2959w, 2927w, 1621m, 1571w, 1456ms, 1367w; mass spectrum (APCI): m/e (% relative intensity) 371 (M+H)’ (80), 328
Daldiniapyrone and Annularins
Pyrones are among the most important heterocyclic structures in medicinal and natural product chemistry, and specifically, 4-hydroxy-2-pyrones (or a-pyrones) can be found in a wide range of medicinally significant natural products "”
Figure 1 Representative Natural Products containing a-pyrone moiety
These natural products range from complex natural products such as orevactaene,*° and arisugasins,°’ to simple ones such as rosellisin®? with its anti- bacterial properties Figure 1 shows some of these natural products that provoked synthetic interest in our group.
More recently, daldiniapyrone'° (5), along with several other metabolites, was isolated from organic extracts of fruit bodies of Daldinia concentria collected in Europe, during an effort to determine chemical constituents of the Ascomycete Daldinia concentrica The structure of 5 was determined by X-ray crystallographic analysis. Although this compound represents an unprecedented metabolite, there are numerous examples of pyran metabolites from fungi'’* in the literature, including species in the
Xylariaceae For instance, “pyrone-3-acetic acid has been reported from a Xylaria species.'!? Unfortunately, daldiniapyrone was not tested for its biological importance, as researchers were more interested in chemotaxonomy.
MeO MeO MeO MeO Me
9: Annularin A 10: Annularin D 11: Annularin E 12: Annularin G 13: Annularin H Figure 2 Structure of annularins A, D, E, G and H
Freshwater fungi have not been extensively investigated as potential sources of new biologically active compounds Annularins'2 B (7), C (6) and F (8) (Figure 1) along with annularins A (9), D (10), E (11), G (12) and H (13) (Figure 2) were isolated from a fresh water fungus Annulatascus triseptatus, as part of an ongoing research project focusing on chemical studies of freshwater fungal species Annulatascus triseptatus is the first member of a new genus in the family Annulatascaceae (Sordariales) This family occurs commonly on submerged woody debris in lotic habitats throughout North America and also in Costa Rica, Scotland, and Venezuela.
Extract from a fermentation cultures of Annulatascus triseptatus showed antifungal and antibacterial activity Annularins A, B, C, and F were found to exhibit activity against Bacillus subtilis, while only annularin C displayed activity against
Staphylococcus aureus All annularins except for annularins D and E, due to sample limitations, were screened for antifungal activity against Aspergillus flavus, but none showed activity in this assay.
The closest known fungal metabolites include cladobotrins, pestalopyrones, and pestalotins,’? which differ from annularins in the length and/or functionality of the side chain, as well as oxidation and methylation patterns Biosynthetic studies support a polyketide origin for members of this class.'** Annularins G and H are envisioned to arise via an analogous biogenetic route, but with a different ring closure step To our knowledge, the fused a-pyrone-furanone system in annularin F has not been encountered previously among natural products and there are no syntheses of daldiniapyrone or annularins that have been reported in the literature.
While intramolecular cyclizations represent a common synthetic entry to 6- substituted-4-hydroxy-2-pyrones 14, it appears that 7-substituted-4-hydroxy-2-pyrones
15, also prevalent among many natural products, can be accessed directly from readily available 6-methyl-4-hydroxy-2-pyrone 14, more commonly known as triacetic acid, via functionalization at the C7 position (Scheme 1) This method should lead to facile syntheses of daldiniapyrone and annularins More importantly, it would provide a platform to propose the key transformation in the orevactaene synthesis featuring an olefination protocol to connect the side chain with the bycyclic ring system Since orevactaene was the natural product of our interest, it drove our focus to this particular transformation Our original retrosynthesis of daldiniapyrone and annularins is shown in Scheme 1. le) le) C7 funct 8)
C7 funct le) iow => oe MeO 4 ` = MeO
4-hydroxy-2-pyrone triacetic acid 5: Daldiniapyrone
Our retrosynthetic analysis revolves around the key intermediate 16 which is a known compound and has been synthesized before.’ Since 16 appeared to be readily available by the known route, it would be easy to make daldiniapyrone and annularins B,
C and F, if there existed a reliable C7 functinalization methodology Moreover this synthesis would be divergent, as four natural products would be made form the single source.
Functionalization at C7 position of 6-methyl-4-hydroxy-2-pyrone turned out to be precedented.*!*"* However, during our pursuit of syntheses of daldiniapyrone and annularins by following literature precedents, we found various inconsistencies The method that was in the process of development at the onset of this project by other members of our research group is shown in Scheme 2 and we decided to utilize it.
2 equiv n-BuLi RI, -78 °C to rt, 12h O
⁄ THF:HMPA (6:1) ZA Li 77-35% overall Zz R
To start our syntheses of daldiniapyrone and annularins and also to test a developing C7 functionalization methodology, we needed a quick and reliable method to come up with multi-gram quantities of synthetic intermediate 16 Despite many known pathways’””°, condensation of malonyl! dichloride (22) with B-keto ester”! 20 still remains to be the more straightforward and economical method for the preparation of 4-hydroxy- 2-pyrone 21 (Scheme 3). ut, Ate ° Ọ 20 CH,Cl,, rt, 7 d
40% Oo” ~OMe ®) benzene im, DMF, TIPSCI
TIPSO O 9 0 ) P2804 pS Gee ewN OMe Cl Cl
Unfortunately, we could not repeat literature yields In our hand, condensation of
20 with malonyl dichloride proceeded to give compound 21 in only 40% yield. Compound 24 was also formed as a result of self-condensation of B-keto ester 20; fortunately for us 21 and 24 were easily separated by simple acid-base extraction By- product 24 was the sole product when the reaction was refluxed in DCM, even the trace amount of desired 21 was not detected upon these conditions Long reaction time, 5 d, and rt were the best conditions for this condensation After 5 d there was still some of B- keto ester 20 left but malonyl dichloride appeared to be all gone and longer reaction time or additional equiv of malonyl dichloride did not improve the outcome Since we needed catalyst such as ZnBr; expedited the process but lowered the yield It was also reported”! that condensation of 23 with malonyl] dichloride affords 21 in 70% yield and decreased the reaction time (Scheme 3) but, in our hands, formation of the desired condensed product 21 proceeded in a very low yield Attempts to use a cheaper starting material such as 25 only produced self-condensed product 24.
2) nBul or Etl, -78 °C to rt, 12h QO” “OMe
27 K2CO3, Me;SO¿ acetone, reflux
1) nBuLi or LDA or KHMDS, fe) ( ÀQ _ THEHMPA -78 °C, 2h (`0
MeO 2) nBul, -78 °C to rt, 16 h MeO
Our next disappointment came when we tried to alkylate compound 21, to access intermediates 26 and 27, as no desired products were detected and only decomposition of the starting material was seen (Scheme 4) Compound 16 was easily obtained upon refluxing 21 in acetone with Me2SO, and K2CO3 Alkylation of 16 was also problematic.Several different bases were tried, such as zBuLl, LDA and KHMDS, but still no alkylated product was observed At this point, as the parallel development of C7 functionalization of 6-methyl-4-hydroxy-2-pyrones ran into problems, our major goal became the syntheses of daldiniapyrone and annularins.
To go around the problematic C7 alkylation we revised our original synthetic plan and decided to condense B-keto esters 28 and 29 with malonyl dichloride using the same conditions as described before to afford products 27 and 26 in moderate yields (Scheme 5) Compound 28 is available commercially and 29 is readily available via an alkylation of a commercially available B-keto esters 20 Self-condensation by-products were also seen.
The synthesis of daldiniapyrone (5) was completed through treating 2-pyrone 26 with methyl sulfate and K2CO; in refluxing acetone The spectroscopic data of synthetic daldiniapyrone were identical to those reported.” Following the same procedure compound 17 was also obtained in 95% yield.
To continue with the synthesis of annularin F we still had to functionalize C7 position of 16 Allylic oxidation'” at C7 of 16 was best accomplished using 5.0 equiv of
SeO; in dioxane at 150 °C in a sealed tube for 2 d (Scheme 6) Under these conditions starting material 16 was completely consumed at the end of the reaction and the only isolatable product was aldehyde 31 Higher temperatures than 150 °C or longer reaction times than 2 d did not improve the yield of this reaction.
O 150 °C, dioxane 19) [O] | 10) ZANT ZN OH |" Cae MeO 46% MeO MeO
Our attempts to improve the outcome of the oxidation reaction and to avoid the toxic SeO;, included bromination of C7 position of 21 with NBS, which would follow by hydrolysis of 34, if successful, and oxygenation using freshly prepared Davis oxaziridine.