Volume 26, Issue 17 e202300193
Review
Open Access

A Historical Review of the Total Synthesis of Natural Products Developed in Portugal

Duarte B. Clemente

Duarte B. Clemente

Centro de Química Estrutural, Institute of Molecular Sciences Faculty of Sciences, University of Lisbon, Campo Grande, 1749-016 Lisbon, Portugal

Search for more papers by this author
Dr. Jaime A. S. Coelho

Corresponding Author

Dr. Jaime A. S. Coelho

Centro de Química Estrutural, Institute of Molecular Sciences Faculty of Sciences, University of Lisbon, Campo Grande, 1749-016 Lisbon, Portugal

Search for more papers by this author
First published: 13 March 2023

Graphical Abstract

This review summarizes the strategies for the natural product synthesis that were developed in Portugal in the past five decades, including alkaloids, cyclitols, phenylpropanoic acids, γ-butyrolactones, fatty alcohols, xanthones and nucleosides.

Abstract

Total synthesis of natural products is an important discipline of organic chemistry that has enabled the development of new synthetic methods and strategies for the preparation and study of the structure and reactivity of complex naturally occurring products. In this review we summarize the synthetic strategies developed in Portugal by several research groups for the synthesis of bioactive natural products including alkaloids, cyclitols, fatty alcohols, phenylpropanoic acids, γ-butyrolactones, xanthones and nucleosides.

1 Introduction

On the occasion of EurJOC's 25th anniversary, we would like to provide an account on the total synthesis of natural products developed in Portugal, being the Portuguese Chemical Society (SPQ) one of EurJOC's founding and owner societies.

Total synthesis is an important field of chemical research that continues to develop into new fields of scientific research and technological development for the benefit of science and society. Several authors have contributed worldwide to the great progress achieved in the total synthesis of natural products over the years.1, 2 Concurrently, over the past fifty years, a few research groups have advanced the synthetic strategies for the synthesis of natural products in Portugal. Historically mainly two research groups have made contributions to this field in Portugal: the group of A. M. Lobo-S. Prabhakar on the total synthesis of several natural alkaloids from 1977 to 2005 and the group of C. D. Maycock on the total synthesis of cyclitols and derivatives from 1992 to 2014. More recently, contributions by the group of A. P. Rauter on the synthesis of nucleosides, M. R. Ventura on the synthesis of phenylpropanoids, C. M. M. Afonso on the synthesis of xanthones, M. M. A. Pereira on the synthesis of indolizidine alkaloids, N. R. Candeias-P. M. P. Gois on the synthesis of viridicatin alkaloids, L. M. Ferreira-P. S. Branco on the synthesis of guanidine alkaloids and C. A. M. Afonso and co-workers on the synthesis of pyrrole-aminoimidazole alkaloids complete the endeavours of synthetic organic chemists to develop total synthesis strategies in Portugal.

This review provides an overview of the total synthesis of natural products developed in Portugal. The strategies are organized based on the natural product class, including one fatty alcohol, ten cyclitols, one nucleoside, two xanthones, and seventeen alkaloids. The discussion of each class is presented in chronological order, with the most recent advances in the total synthesis of alkaloids being summarized last. There are several total syntheses for most of these natural products, however, for the purpose of this review, we will not compare key methodologies or retrosynthetic strategies. Instead, the total syntheses will be presented in a forward sense. Nonetheless, a comprehensive list of reported protocols for each natural product is included as Supporting Information. The scope of this review will not cover reports on hemisynthetic strategies or synthesis of non-natural products.

2 Total Synthesis of Fatty Alcohols

2.1 Maycock's total synthesis of (+)-sulcatol

Afonso et al. reported the synthesis of both (+)-sulcatol and (−)-sulcatol (not discussed) in 1989 (Scheme 1).3 The synthesis started with the biocatalysed reduction of ketone 1 with Baker's yeast to yield carbinol 2 as a mixture of diastereoisomers. Ester hydrolysis followed by acetylation gave acid 3, which was submitted to the Barton's decarboxylation (reaction conditions not reported) followed by hydrolysis of the acetyl group to afford (+)-sulcatol in 16 % overall yield over five steps. In addition to this protocol, there are thirteen other asymmetric total syntheses of (+)-sulcatol reported by several authors from 1975 to 2001.4

Details are in the caption following the image

Maycock's total synthesis of (+)-sulcatol (1989).3

3 Total Synthesis of Cyclitols and Derivatives

Preparation of natural products derived from quinic acid, a natural product first isolated from species of the Cinchona genus, has been a major focus in the field of total synthesis in Portugal. Contributions of the Maycock research group to this field resulted in the development of several synthetic routes for the preparation of ten natural products (Figure 1) and several unnatural derivatives and analogues (not discussed).

Details are in the caption following the image

Natural cyclitols and derivatives synthesised in Portugal.

3.1 Maycock's asymmetric total synthesis of (+)-negamycin

The Maycock group's earliest report of a total synthesis is on the asymmetric preparation of peptide-like antibiotic (+)-negamycin in 1992 (Scheme 2). Starting with (−)-quinic acid, triol 4 was obtained by a reported synthetic sequence, involving protection of the 1,2-cis diol, hydroxyl group protection and reduction with lithium aluminium hydride.6 Oxidative cleavage of 4 with sodium periodate gave ketone 5. Benzoylation of 5 followed by ketone reduction with sodium borohydride afforded the corresponding alcohol as a mixture of diastereomers R,R,S,R-6 and R,R,R,R-6. The major isomer R,R,R,R-6, bearing the desired configuration, was benzoylated to give 7, whereas the minor R,R,S,R-6 was converted into 7 via a Mitsunobu reaction. Removal of the isopropylidene group with 1,2-ethanedithiol and boron trifluoride diethyl etherate, followed by oxidative ring opening of the formed diol and immediate reduction of the formed aldehyde generated 1,6-diol 8. Treatment of 8 with Hünig's base resulted in a 1,4-benzoyl migration, leading to 1,5-diol 9, whose primary and secondary hydroxyl groups were protected, respectively, with tert-butyldimethylsilyl chloride and benzyloxymethyl chloride, to afford 10. Hydrolysis of the benzoate esters, followed by conversion of the resulting 1,4-diol to the dimesylate and consequent displacement of the mesylates with sodium azide afforded 11. Reduction of the azido groups into amino groups followed by protection with tert-butyloxycarbonyl anhydride generated 12. Preparation of the acid 13 was achieved by removal of the silyl group with tetra-n-butylammonium fluoride and oxidation with sodium periodate. Reaction of 13 with ethyl chloroformate and N-methylhydrazinoacetate (14) generated hydrazide 15. Finally, deprotection of the BOM-protected alcohol by hydrogenation, the Boc-amino group by hydrolysis and hydrolysis of the ester group by treatment with trifluoroacetic acid yielded the trifluoroacetate salt of (+)-negamycin, in up to 14 % overall yield over twenty-three steps.5 This protocol is part of a series of 19 total syntheses reported by several authors from 1972 to 2014 (including 15 asymmetric strategies).4

Details are in the caption following the image

Maycock's total synthesis of (+)-negamycin TFA salt (1992).5

3.2 Maycock's asymmetric total syntheses of (+)-eutypoxide B, (+)-epoformin, (+)-harveynone, (−)-asperpentyn, (+)-epiepoformin, (−)-theobroxide, (+)-bromoxone, and (−)-LL-C10037α

Following the preparation of (+)-negamycin, the Maycock group reported, from 1997 to 2014, the asymmetric synthesis of several polyoxygenated cyclohexanone natural products bearing an epoxide functionality: (+)-eutypoxide B,7 (+)-epoformin,8 (+)-harveynone,9 (−)-asperpentyn,9 (+)-epiepoformin,9 (−)-theobroxide,9 (+)-bromoxone10 and (−)-LL−C10037α.11

In 1997 Barros et al. reported an enantioselective total synthesis of (+)-eutypoxide B, a metabolite of the Eutypa lata fungus (Scheme 3). Starting with 5, conjugated ketone 16 was prepared by dehydration. Conjugate addition of the appropriate mixed cuprate reagent to 16 afforded ketone 17, and its corresponding base-promoted elimination product 18. Conversion of the remaining 17 to 18 was achieved with sodium hydroxide. Protection of the alcohol in 18 as silyl ether afforded 19. Epoxidation of 19 with tert-butyl hydroperoxide generated 20, from which Luche reduction afforded 21. Subsequent silylation and ozonolysis afforded aldehyde 22, which underwent 1,2-addition of (2-methyl-1-propen-3-yl)magnesium chloride (23) to form alcohol 24 as a mixture of diastereomers. Finally, oxidation with Dess-Martin periodinane, base-catalysed double bond isomerization and cleavage of the silyl ethers with tetra-n-butylammonium fluoride afforded (+)-eutypoxide B in 22 % overall yield over twelve steps.7 This protocol is one of the three strategies reported by different authors for the total synthesis of this natural product.4

Details are in the caption following the image

Maycock's total synthesis of (+)-eutypoxide B (1997).7

The first enantioselective total synthesis4 of Penicillium claviform metabolite (+)-epoformin was reported by Barros et al. in 1999 (Scheme 4). The designed route started with the protected quinic acid derivative 25, in which benzoylation of the primary hydroxyl group and subsequent oxidation of the secondary alcohol yielded ketone 26. 1,2-Addition of methylcerium chloride to 26 generated triol 27 as a mixture of diastereomers, from which ketone 28 was obtained by oxidative cleavage with sodium periodate. Base-catalysed elimination and removal of the isopropylidene group selectively afforded 29, which on secondary hydroxyl group protection followed by epoxidation afforded 30 selectively. Dehydration upon treatment with triflic anhydride and Hünig's base yielded enone 31. Selective reduction of the keto group in 31 with L-selectride® followed by acetylation with acetic anhydride led to 32. Finally, cleavage of TBS protecting group followed by Dess–Martin oxidation and hydrolysis of acetyl protecting group gave (+)-epoformin in 12 % overall yield over thirteen steps.12

Details are in the caption following the image

Maycock's total synthesis of (+)-epoformin (1999).8

In 2000 Maycock and co-workers reported a divergent strategy starting with quinic acid derivative 5 that allowed the asymmetric synthesis of four natural cyclitols isolated from fungus: (+)-harveynone, (−)-asperpentyn, (+)-epiepoformin and (−)-theobroxide (Scheme 5). Thus, the starting material was silylated to yield protected alcohol 33, which upon treatment with sodium hydroxide generated an isomeric mixture of hydroxy ketones including desired 34. Epoxidation of 34 with hydrogen peroxide afforded a mixture of isomeric epoxides including desired 35. Treatment of 35 with acetic anhydride led to a mixture of enones including desired 36. Iodination of 36 gave key intermediate iodoenone 37. Introduction of an acetylated side chain was achieved via a Stille coupling, leading to 38, from which (+)-harveynone was secured by cleavage of the silyl ether protecting group. The synthesis afforded (+)-harveynone in 7 % overall yield over seven steps. Luche reduction of the keto group in 38 followed by deprotection gave (−)-asperpentyn. This route constituted the first reported synthesis4 of (−)-asperpentyn, achieved in 5 % overall yield over eight steps. The remaining target molecules mentioned above were prepared from 37, which underwent a α-methylation via Stille coupling to yield 39. Desylilation of 39 afforded (+)-epiepoformin, which was obtained from 5 in 7 % overall yield over seven steps. Finally, Luche reduction of the keto group in 39 followed by removal of the protecting groups originated (−)-theobroxide in 5 % overall yield over eight steps.9

Details are in the caption following the image

Maycock's total syntheses of (+)-harveynone, (−)-asperpentyn, (+)-epiepoformin and (−)-theobroxide (2000).9

In 2003 Barros et al. reported a total synthesis of (+)-bromoxone, a natural product isolated from species of the Ptychodera genus, employing aziridines as simultaneous protecting and directing group for the stereocontrolled total synthesis (Scheme 6). The synthesis started with 16, which afforded iodoenone 40 by iodination. Installation of the aziridine moiety was achieved by reaction with 4-methoxybenzylamine (41) via Michael addition followed by cyclization reaction to yield 42. Treatment of 42 in basic media followed by silylation led to protected alcohol 43. The key epoxidation directed by the aziridine group afforded 44 as a single diastereomer. Finally, selective aziridine ring-opening and bromination followed by de silylation furnished (+)-bromoxone in 25 % overall yield over seven steps.10 This protocol is part of a collection of eight asymmetric total synthesis of (+)-bromoxone reported by several authors from 1994 to 2012.4

Details are in the caption following the image

Maycock's total synthesis of (+)-bromoxone (2003).10

The asymmetric total synthesis of bacterial metabolite (−)-LL-C10037α, reported by Maycock et al. in 2014, was achieved through the same directing group strategy previously reported in 200310 by the same group (Scheme 7). Thus, resolution of the racemic iodoenone 45 by aziridination with (R)-(+)-4-methoxy-α-methylbenzylamine (46) afforded cis isomers R,R,S,R-47 and S,S,R,R-47. Desired ketone R,R,S,R-47 was converted to enone 49 by enolization with lithium diisopropylamide followed by reaction with tert-butyl phenylsulfinimidoyl chloride (48). Epoxidation with hydrogen peroxide afforded 50, which was treated with hydrazoic acid to give 51 upon aziridine ring cleavage-elimination. Staudinger reduction of the azide moiety in the presence of acetic anhydride followed by deprotection of the hydroxyl group gave (−)-LL-C10037α. This synthetic sequence furnished (−)-LL-C10037α in 19 % overall yield over 6 steps.11 The protocol outlined here is the most recent total synthesis of (−)-LL-C10037α in addition to seven other total synthesis of this natural product.4

Details are in the caption following the image

Maycock's total synthesis of (−)-LL-C10037α (2014).11

3.3 Maycock's total synthesis of (−)-methyl shikimate

In 1999 Alves et al. reported the synthesis of (−)-methyl shikimate together with unnatural (−)-3-epi-methyl shikimate (not discussed). The synthesis started with 52, which was obtained from (−)-quinic acid by selective protection of the vicinal trans diol in the form of tartrate derivatives (Scheme 8). Swern oxidation of 52 led to a mixture of ketones, i. e., ketone 53 and the desired enone 54. Conversion of the remaining 53 into 54 was achieved by dehydration. Reduction of 54 with L-selectride® followed by deprotection of the 1,2-diol furnished (−)-methyl shikimate in 56 % overall yield over four steps.13 This strategy is part of a collection of fourteen reported total synthesis of this natural product by several authors from 1981 to 2009 (ten asymmetric protocols).4

Details are in the caption following the image

Maycock's total synthesis of (−)-methyl shikimate (1999).13

4 Total Synthesis of γ-Butyrolactones

4.1 Maycocks's total synthesis of (+)-nephrosteranic acid

In 2003 Barros et al. reported the synthesis of (+)-nephrosteranic acid, a C-4 methyl containing paraconic acid (Scheme 9). The synthesis started with bis-acetal dioxane 56, prepared from L-(+)-tartaric acid 2,2,3,3-tetramethoxybutane (55).14 The core lactone was formed by conversion of 56 into the corresponding dithioester, followed by alkylation using lithium diisopropylamide and then addition of dodecanal to give lactone 57. Hydrolysis of the acetal and the thioester (transesterification) gave 58, which underwent mesylation and elimination to afford butenolide 59. Hydrogenation of 59 gave a mixture of diastereomers favouring cis-60, which resulted in a final mixture favouring the trans isomer (1 : 4.6) upon treatment with DBU. Finally, hydrolysis of the ethyl ester and α-methylation furnished (+)-nephrosteranic acid in 43 % overall yield over nine steps.15 This protocol is part of a series of 18 total syntheses reported by several authors from 1994 to 2020.4

Details are in the caption following the image

Maycock's total synthesis of (+)-nephrosteranic acid (2003).15

4.2 Maycocks's total synthesis of (−)-muricatacin

In 2009 Barros et al. reported the synthesis of (−)-muricatacin, a biologically active annonaceous acetogenin. Monodihydroxylation of diene 61 followed by oxidative cleavage using sodium periodate gave aldehyde 62 (Scheme 10). Witting olefination of 62 with in situ generated n-undecyltriphenylphosphorane (from 63) afforded diene 64 (Z configuration at the newly formed bond). Finally, hydrogenation of the C−C double bonds followed by TBS-removal afforded (−)-muricatacin in 37 % overall yield over five steps from 61.16 This protocol is part of a series of 48 total syntheses reported by several authors from 1991 to 2017 (including both isomers).4

Details are in the caption following the image

Maycock's total synthesis of (−)-muricatacin (2009).16

5 Total Synthesis of Phenylpropanoic Acids

5.1 Ventura's total synthesis of (+)-piscidic acid and cimicifugic acid L

In 2015, Miranda et al. reported the syntheses of (+)-piscidic acid and its 3,4-dimethoxycinnamic acid ester derivative (cimicifugic acid L). The synthesis started with bis-acetal dioxane 65 using similar strategy used for the preparation of 56 but using D-(−)-tartaric acid (not shown). Dithioester 65 was alkylated using lithium diisopropylamide and 4-benzyloxybenzyl bromide (66) to afford 67 (Scheme 11). Hydrolysis of the acetal and the thioester (transesterification) gave 68. From this common intermediate, catalytic hydrogenation followed by methyl esters hydrolysis gave (+)-piscidic acid in 56 % overall yield over five steps from 65.17 This protocol is part of a series of 8 total syntheses reported by several authors from 1966 to 2015.4 Esterification of alcohol 68 with 3,4-dimethoxycinnamoyl chloride (69) resulted in the formation of cinnamate ester 70, which after Iron(III)-catalysed removal of the benzyl group and hydrolysis of methyl esters using LiI gave cimicifugic acid L as a mixture of trans/cis isomers in 21 % overall yield over six steps from 65.17 This is the first total synthesis reported for this natural product.4

Details are in the caption following the image

Ventura's total syntheses of (+)-piscidic acid and cimicifugic acid L (2015).17

6 Total Synthesis of Nucleosides

6.1 Rauter, Sinaÿ and Blériot's total synthesis of the miharamycin B core

In 2008 Marcelo et al. reported the first synthesis of protected myharamycin B, a nucleoside with antibacterial activity (Scheme 12). The synthesis started with the construction of the bicyclic carbohydrate core. Treatment of D-glucose derivative 71 with samarium(II) iodide afforded alkene 72 through a 5-exo-dig ketyl-alkyne cyclisation. Ozonolysis of 72, with subsequent ketone reduction and benzylation yielded benzylidene-protected glucoside 73. Regioselective ring opening of the benzylidene acetal protecting group by lithium aluminium hydride-aluminium chloride generated 74 followed by Swern oxidation to give the corresponding aldehyde, which was directly alkylated with vinylmagnesium bromide, affording a diastereomeric mixture of allylic alcohols containing the desired alcohol 75. Ozonolysis of 75 and further Pinnick oxidation, followed by esterification with methyl iodide yielded methyl ester 76. Deoxyazidation of the secondary alcohol of 76 (with inversion of configuration) gave azide 77, followed by acetolysis to yield anomeric acetate 78. Finally, N-glycosylation of purine 79 with glycosyl donor 78 yielded the desired N9-nucleoside 80 as the major product, whose azido group was hydrogenated, leading to 81, from which the protected miharamycin B was obtained by peptide coupling with 82. This synthetic sequence afforded protected myharamycin B in 2 % yield over sixteen steps.18 The first total synthesis of myharamycin B was only recently reported by others.4

Details are in the caption following the image

Rauter, Sinaÿ and Blériot's total synthesis of protected miharamycin B (2008).18

7 Total Synthesis of Xanthones

7.1 Afonso's total syntheses of yicathins C and B

The total synthesis of xanthones has been explored by the group of Carlos M. M. Afonso. In 2020 Loureiro, Magalhães, Soares et al. reported the total synthesis of yicathins C and B, two marine natural products isolated from a fungus of the Aspergillus genus that exhibit antibacterial and antifungal activities (Scheme 13). The synthesis started with 4-bromo-3,5-dimethoxybenzoic acid (83), which underwent reduction to benzylic alcohol with borane tetrahydrofuran complex followed by silylation to yield protected alcohol 84. Lithium-halogen exchange using nBuLi generated the corresponding organolithium reagent that was treated with aldehyde 85 (prepared from orcinol in two steps and 86 % yield) to yield alcohol 86. Dess-Martin oxidation followed by removal of protective groups in acidic conditions furnished 87, from which 88 was secured via a microwave-assisted intramolecular nucleophilic aromatic substitution reaction. Finally, Jones oxidation of 88 gave yicathin C in 8 % yield over seven steps. In addition, yicathin B was obtained by Fischer esterification, in 4 % yield over eight steps.19 The total synthesis described above is the only protocol reported to date for the preparation of these xanthones natural products.4

Details are in the caption following the image

Afonso's synthesis of yicathins C and B (2020).19

8 Total Synthesis of Alkaloids

The development of synthetic routes for the total synthesis of alkaloids in Portugal has been historically mostly achieved by the Lobo and Prabhakar research group (Figure 2). Furthermore, considering the natural products presented herein, this research group developed the first total synthesis reported in Portugal - the synthesis of (±)-cis-alpinigenine published in 197720 More recently, Pereira, Candeias, Góis, Ferreira, Branco, Siopa, Gomes, Afonso and others have developed methodologies for the synthesis of natural alkaloids.

Details are in the caption following the image

Natural alkaloids synthesised in Portugal.

8.1 Lobo and Prabhakar's total syntheses of isoquinoline alkaloids (±)-cis-alpinigenine and (±)-alpinigenine

In 1977 Lobo, Prabhakar and co-workers reported the total synthesis of (±)-cis-alpinigenine, a tetrahydroisoquinoline alkaloid found in the species of the Papaver genus (Scheme 14).20 Later, in 1981, the same research group reported that (±)-alpinigenine was also produced as a minor product of the same synthesis.21 The developed route started with tetracyclic base 89, prepared from homoveratrylamine and o-homoveratric acid in four steps (72 % yield, not shown).22 This on N-methylation followed by treatment with HCl afforded methochloride 90. Hofmann elimination promoted by Amberlite® IRA-400 gave tricyclic olefin 91. Halogenation of the alkene followed by solvolysis of the resulting halohydrin led to diol 92. Oxidation of 92 with periodic acid cleaved the vicinal diol into dialdehyde 93. The synthesis was completed by a photo-induced ring-closure reaction via intermediate dienol 94 that underwent an intramolecular hetero-Diels-Alder reaction. The endo-cyclisation product (±)-cis-alpinigenine was obtained in 13 % yield from 89, and the thermodynamically less stable exo-addition product (±)-alpinigenine was obtained in 1 % as a minor product.20, 21 To the best of our knowledge, there are no asymmetric total syntheses of these natural products.4

Details are in the caption following the image

Lobo and Prabhakar's total syntheses of (±)-cis-alpinigenine and (±)-alpinigenine (1977).20, 21

8.2 Lobo and Prabhakar's total syntheses of ismine, and pyrrolophenantridine alkaloids hippadine, pratosine, vasconine, assoanine and oxoassoanine

From 1985 to 1997, the Lobo and Prabhakar group developed several synthetic routes for the preparation of various alkaloids isolated from species of the Amaryllidacaea genus: ismine,23, 24 hippadine,25 pratosine,24, 26 vasconine,24 assoanine,24 and oxoassoanine.24 With the exception of ismine, these alkaloids share a pyrrolophenantridine skeleton.

In 1985 Lobo, Prabhakar and co-workers reported a synthetic route to access ismine starting with hydroxamic acid 95 (Scheme 15), prepared from 2-bromo-4,5-methylenedioxybenzoyl chloride and N-phenylhydroxylamine (not shown). Conversion of 95 into its corresponding boron difluoride complex using boron trifluoride diethyl etherate followed by a photo-induced cyclization afforded boron complex 96, which was hydrolyzed to give cyclic hydroxamic acid 97. Oxidative ring-opening with sodium periodate afforded nitro-carboxylic acid 98, which on treatment with sodium borohydride and titanium tetrachloride was reduced to give N-norismine (99). Finally, methylation of 99 completed the synthesis of ismine in 20 % overall yield.23 Later, in 1997, the same research group reported an alternative approach for the formal synthesis of ismine starting with the condensation of 2-bromo-4,5-methylenedioxybenzyl chloride (100) with 2-nitrophenol to yield benzylphenylether 101. Then Pd/C-catalysed hydrogenation of the nitro group followed by radical cyclisation with tributyltin hydride generated N-norismine (99) in 13 % yield over three steps.24

Details are in the caption following the image

Lobo and Prabhakar's total syntheses of ismine (1985, 1997).23, 24

In 1987 Lobo, Prabhakar and co-workers reported the first total synthesis of hippadine in two steps starting with the previously prepared hydroxamic acid 97 (Scheme 16). The synthetic route involved a base-catalysed conjugate addition with methyl propiolate to afford the compound 102, followed by a quadruple cascade reaction involving 1-aza-1’oxa-[3,3]-sigmatropic rearrangement, ester hydrolysis, decarboxylation, and dehydration, to give hippadine in 17 % yield over two steps.25 In addition to this protocol, there are eighteen other total synthesis reported by others from 1987 to 2014.4

Details are in the caption following the image

Lobo and Prabhakar's total synthesis of hippadine (1996).25

In 1996 Lobo, Prabhakar and co-workers reported the total synthesis of pratosine, for which the pyrrolophenantridine scaffold was constructed similarly to its methylenedioxy analogue hippadine (Scheme 17). Thus, hydroxamic acid 103, prepared from 2-bromo-4,5-dimethoxybenzoyl chloride and N-phenylhydroxylamine (not shown), complexed with boron trifluoride to generate boron difluoride complex 104, which underwent a photo-induced cyclisation followed by hydrolysis to afford 105. Conjugate addition of 105 to methyl propiolate afforded enol ether 106. Finally, a quadruple cascade reaction led to the formation of pratosine in 60 % yield from 103.26 In addition to this protocol, there are ten other protocols reported by others from 1994 to 2016.4

Details are in the caption following the image

Lobo and Prabhakar's total synthesis of pratosine (1996).26

In 1997 Lobo, Prabhakar and co-workers reported a collective total synthesis of several Amaryllidacaea alkaloids, in which vasconine was employed as the common intermediate for the divergent syntheses of assoanine, oxoassoanine and pratosine (Scheme 18). The synthesis started with condensation of 2-bromo-4,5-dimethoxybenzaldehyde (107) with 2-β-hydroxyethylaniline (108) to afford an aldimine, which was reduced to aminoalcohol 109. Radical cyclisation reaction of 109 upon treatment with tributyltin hydride generated 110, followed by ring closure mediated by phosphorous tribromide, afforded vasconine in 7 % overall yield. The synthesis of assoanine was completed by reduction of the iminium ion to afford amine in 6 % overall yield. Finally, oxoassoanine was synthesized by oxidation of vasconine with alkaline hydrogen peroxide to provide lactam oxoassoanine in 4 % overall yield, whereas pratosine was synthesized by oxidation with potassium ferricyanide to give pratosine as the minor product in 2 % overall yield and oxoassoanine as the major product.24 It is important to note that several total syntheses of these four natural products were reported in the last two decades demonstrating the relevance of this work.4

Details are in the caption following the image

Lobo and Prabhakar's total synthesis of vasconine and divergent syntheses of assoanine, oxoassoanine and pratosine (1997).24

8.3 Lobo and Prabhakar's total syntheses of indolocarbazole alkaloids arcyriaflavin A and staurosporinone

Natural products bearing indolocarbazole scaffolds have drawn considerable attention to synthetic chemists since they exhibit remarkable biological activities. Particularly, Lobo, Prabhakar and co-workers’ contributions to this topic consisted in the development of two synthetic strategies for the total synthesis of arcyriaflavin A, in 199527 and 2000,28 and a total synthesis of staurosporinone in 2003.29

Lobo, Prabhakar and co-workers reported results of their first approach to the preparation of arcyriaflavin A, which consisted in a two-step synthesis (Scheme 19). It started with the condensation of 3-mercaptoindole (111) with 3,4-dichloromaleimide (112) to yield bis-sulfide 113. Stoichiometric Pd-mediated oxidative coupling of the indole moieties, [4+2] cycloaddition and sulfur extrusion afforded arcyriaflavin A in 9 % overall yield. An improved synthesis, reported in 2000, consisted of condensation of 1,2-dithiine 114 with 3,4-dibromomaleimide (115) to generate bis-sulfide 116, followed by an intramolecular Diels-Alder reaction and consequent double retro-cheletropic sulfur extrusion to yield arcyriaflavin A, in 34 % overall yield.27, 28

Details are in the caption following the image

Lobo and Prabhakar's total syntheses of arcyriaflavin A (1995, 2000).27, 28

The synthesis of staurosporinone was achieved through a two-step synthesis (Scheme 20). Coupling of 2,2’-biindole (117) with tetramic acid derivative 118 afforded intermediate 119, which underwent oxidative photocyclization to give staurosporinone in 46 % overall yield.29

Details are in the caption following the image

Lobo and Prabhakar's total synthesis of staurosporinone (2003).29

8.4 Lobo and Prabhakar's total synthesis of pyrroloindole alkaloids (±)-desoxyeseroline and asymmetric total synthesis of (−)-debromoflustramine B

From 1995 to 2007, the Lobo and Prabhakar's group developed several synthetic routes for the preparation of pyrroloindole alkaloids. They reported three racemic syntheses of (±)-desoxyeseroline, in 1995,30 200131 and 2005,32 the first asymmetric total synthesis of (−)-debromoflustramine B, in 2001,33 as well as an alternative approach to one of the steps in the synthesis of this alkaloid, in 2007. Additionally, the asymmetric total syntheses of unnatural (+)-debromoflustramine B and (+)-debromoflustramide B were reported.33

Their first approach to the synthesis of racemic desoxyeseroline initiated with N-methylaminoskatole (120) (Scheme 21). Conjugate addition of 120 to methyl propiolate generated enaminoester 121, which underwent a thermo-induced [3,3]-sigmatropic rearrangement and cyclisation in diphenyl ether to afford tricycle 122 in 51 % yield. Acylation of 122 followed by hydrogenation yielded methyl ester 123 as a mixture of diastereomers. Saponification and subsequent Barton's decarboxylation provided carbamate 124 in 51 % yield. Reduction of 124 with lithium aluminium hydride led to amine 125, from which (±)-desoxyeseroline was accessed by reductive methylation with sodium cyanoborohydride and formaldehyde. Overall, the synthesis of this alkaloid was achieved in 9 % yield over eight steps.30

Details are in the caption following the image

Lobo and Prabhakar's total syntheses of (±)-desoxyeseroline (1995, 2005).30, 32 ODCB, 1,2-dichlorobenzene.

Later in 2005, a novel preparation of this alkaloid in higher yield was reported by using a similar synthetic route, differing in two steps (Scheme 21): improvement of the cyclisation step was achieved by performing the reaction in o-dichlorobenzene (ODCB), which led to an improved yield of 91 %; and decarboxylation of 123 was achieved via a benzophenone oxime ester intermediate, which afforded 124 in 69 % yield. Consequently, the novel synthesis afforded (±)-desoxyeseroline in 21 % overall yield.32

In 2001, the Lobo and Prabhakar's group reported a synthesis of this natural product through a different approach for the construction of the pyrroloindole skeleton of (±)-desoxyeseroline, starting with N-phenylhydroxylamine (126) (Scheme 22). Chemoselective N-acylation of 126 with methyl chloroformate afforded the corresponding hydroxamic acid, which underwent O-acylation with 2-phenylsulfanylpropanoic acid to give 127. Treatment of 127 with potassium hexamethyldisilazane gave ortho-aminophenylacetic ester 128 (and its para isomer as a minor product) which afforded oxindole 129 upon dehydration. Desulfurization of 129 with tributyltin hydride generated oxindole 130. Conjugate addition of 130 to nitroethylene led to β-nitroethyloxindole 131 and subsequent hydrogenation afforded pyrrolidone 132. This intermediate was treated with an excess of methyl iodide in the presence of sodium hydride to yield a mixture of isomers 133 and 134 that furnished oxindole 135 upon acid hydrolysis. Completion of the synthesis was achieved by reduction of 135, yielding (±)-desoxyeseroline in 11 % overall yield.31

Details are in the caption following the image

Lobo and Prabhakar's total synthesis of (±)-desoxyeseroline (2001).31

A different approach to build the pyrroloindole core was investigated by Lobo, Prabhakar and co-workers in the total synthesis of (−)-debromoflustramine B. The authors reported the first asymmetric synthesis of this alkaloid in 2001, which started with N-acetyl-ʟ-tryptophan methyl ester (135) containing already all the carbon atoms of the pyrroloindole core in its structure (Scheme 23). Cyclisation of 135 promoted by tert-butyl hypochlorite led to pyrroloindole 136, with subsequent C-prenylation using 137, yielding 138 as a mixture of diastereomers. Reduction of the mixture generated amine exo-139, which was converted to the corresponding C,N-diprenyl intermediate exo-140. A decrease in the diastereomeric ratio was observed throughout these reaction steps, favouring the formation of the undesired endo product given its intrinsic higher stability opposed to the exo isomer. Saponification followed by reaction with Barton's reagent afforded a mixture of Barton's esters from which the minor ester exo-141 was isolated. Treatment with tris(phenylthio)antimony and air exposure promoted a decarboxylative hydroxylation to generate alcohol 142. Dehydration of 142 followed by deacetylation generated imine 144 that was reduced to furnish secondary amine 145, which was readily methylated, yielding (−)-debromoflustramide B. The eleven-step synthetic route afforded this natural product in 0.1 % overall yield.33

Details are in the caption following the image

Lobo and Prabhakar's first total synthesis of (−)-debromoflustramine B (2001).33

An analogous route led to the preparation of the non-occurring (+)-debromoflustramine B from endo-141 in 0.8 % overall yield. Additionally, (+)-debromoflustramide B was prepared from endo-141 in 14 % yield over two steps (2 % from 136).33 Later in 2007, a modification of the C-prenylation step using phase transfer conditions induced diastereoselectivity in this step, favouring the formation of R,S-138 (1 : 2 diastereomeric ratio), thus constituting an improvement in the asymmetric synthesis of (+)-debromoflustramine B and (+)-debromoflustramide B.34

8.5 Pereira's total synthesis of (−)-indolizidine 167B

In 2002, Corvo et al. reported an asymmetric synthesis of (−)-indolizidine 167B from (±)-norvaline (Scheme 24). The synthesis started with a Paal–Knorr pyrrole condensation of 2,5-dimethoxytetrahydrofuran (DMTHF) with the formed (±)-norvaline methyl ester, which after hydrolysis gave carboxylic acid 145. Esterification of 145 with Barton's reagent gave thiohydroxamate ester 146, which was added to chiral acrylamide 147 under irradiation, in which the desired diastereomer 148 was isolated by chromatography as a mixture of diastereomers at C4. Then, 148 underwent an intramolecular Friedel-Crafts acylation to give 149 with retention of configuration upon treatment with boron tribromide. Removal of the thiopyridyl group with nickel(II) chloride and sodium borohydride gave bicyclic keto pyrrole 150, which was hydrogenated to give (−)-indolizidine 167B in 6 % overall yield after 8 steps.35 The authors also reported that enantiomer (+)-indolizidine 167B was obtained from 4S-148. This protocol is part of a series of 41 total syntheses reported by several authors from 1988 to 2020.4

Details are in the caption following the image

Pereiras’ total synthesis of (−)-indolizidine 167B alkaloid (2002).35 DMTHF, 2,5 dimethoxytetrahydrofuran.

8.6 Lobo and Prabhakar's total synthesis of thiazole tryptamide alkaloid bacillamide

In 2005 Lobo, Prabhakar and co-workers reported the first total synthesis of bacillamide, a bacterial algicide containing a thiazole tryptamide scaffold (Scheme 25). The synthesis started with 4-methylthiazole (151), which underwent lithiation followed by treatment with ethyl acetate to afford the C-acetylated thiazole 152. Reaction of 152 with N-bromosuccinimide secured dibromide 153, which was converted into aldehyde 154 by solvolysis in the presence of silver(I) acetate. Consequent Pinnick oxidation afforded carboxylic acid 155, which was coupled with tryptamine to generate bacillamide. The five-step synthesis afforded this alkaloid in 16 % overall yield.36

Details are in the caption following the image

Lobo and Prabhakar's total synthesis of bacillamide.36

8.7 Candeias and Góis’ total synthesis of viridicatin

In 2013, Paterna et al. reported the total synthesis of viridicatin, a quinoline alkaloid and fungal metabolite isolated from several Penicillium species (Scheme 26). The synthesis started with a dirhodium-catalysed ring expansion of diazo compound 157, prepared from the reaction of isatin (156) and ethyl diazoacetate. The authors demonstrated that these reactions can be performed in a one pot fashion with slightly lower yield. Decarboxylation of 158 followed by bromination using N-bromosuccinimide gave 160, which underwent Suzuki-Miyaura cross-coupling with phenylboronic acid to yield viridicatin in up to 64 % overall yield over 5 steps.37 In addition, the authors also synthesised different viridicatin alkaloid derivatives using different boronic acid partners in the cross-coupling reaction. This protocol is part of a series of ten total syntheses reported by several authors from 1954 to 2021.4

Details are in the caption following the image

Candeias and Góis’ total synthesis of viridicatin alkaloid (2013).37

8.8 Ferreira and Branco's total synthesis of (±)-cernumidine

In 2020, Rippel et al. reported the first and only one reported to date total synthesis of (±)-cernumidine, an alkaloid isolated in 2011. The two steps synthesis was initiated by formation of carbamimidoyl-ʟ-proline 161 from the reaction of ʟ-proline and cyanamide (Scheme 27). Then, an oxidative decarboxylation followed by intermolecular trapping of the iminium intermediate by acetylated isoferulic amide (162) gave the natural product in up to 22 % in its racemic form.38

Details are in the caption following the image

Ferreira and Branco's total synthesis of (±)-cernumidine.38

8.9 Afonso and co-workers’ total syntheses of pyrrole-2-aminoimidazole alkaloid agelastatin A

In 2022, Afonso and co-workers have reported two distinct approaches for the construction of the central all-carbon, five-membered ring of agelastatin A, a marine alkaloid isolated from Agelas dendromorpha containing also a pyrrole-2-aminoimidazole core. The relevance of this natural product is highlighted by the twenty-two total syntheses reported since 1999.4

Gomes, Vale et al. reported a total synthesis of (±)-agelastatin A (Scheme 28). The protocol started with furfural, which underwent condensation followed by furan ring-opening in the presence of dibenzylamine and diallylamine to form the mixed Stenhouse salt 163 that in turn underwent a stereoselective 4π-electrocyclization to give the key mixed trans-4,5-diaminocyclopenten-2-one 164. Reduction of the ketone group of 164 followed by protection of the resulting alcohol with tert-butyldimethylsilyl chloride generated 165 stereoselectively. Elaboration of this core was initiated by a Pd(0)-catalysed deprotection of the allyl amine using 1,3-dimethylbarbituric acid (NDMBA) followed by the coupling of the resulting amine with pyrrole-2-carbonyl chloride (166) to yield amide 167. Next, deprotection of the silyl ether with tetra-n-butylammonium fluoride followed by oxidation with 2-iodoxybenzoic acid (IBX) gave cyclopentenone 168. Further elaboration of this core was performed by deprotection of the dibenzylamine using a one pot procedure. Thus, treating 168 with caesium carbonate followed by hydrogenation using Pd/C, followed by coupling of the resulting amine with N-methyl carbamoylimidazole (169) and concomitant cyclization gave urea 170. Finally, bromination of the pyrrole moiety of 170 with dibromantin generated (±)-agelastatin A in 22 % overall yield over eight steps.39

Details are in the caption following the image

Gomes and Afonso's total synthesis of (±)-agelastatin A (2022).39 NDMBA, 1,3-dimethylbarbituric acid.

Vale, Fortunato et al. reported an asymmetric total synthesis of (−)-agelastatin A (Scheme 29). Their approach to the synthesis of this natural product initiated with the conversion of N-allylpyridinium bromide 172 (prepared from pyridine (171) by alkylation with allyl bromide) to the racemic bicyclic aziridine (±)-173 under photochemical continuous flow conditions. Enzymatic kinetic resolution of the racemic mixture was achieved in flow, using immobilized Novozym® 435 CALB lipase and vinyl acetate as an acylating agent, which allowed for the enantioselective acetylation of the undesired R,R,R-173. The resolution furnished the desired enantiomer S,S,S-173 in 93 % enantiomeric excess. Aziridine ring opening with trimethylsilyl azide led to 174 and was followed by urea formation with N-methylcarbamoylimidazole (169), leading to 175. Azide reduction followed by coupling of the aminoalcohol intermediate with pyrrolyl chloride (166) yielded 176. Ley-Griffith oxidation with subsequent cyclization by urea addition to the generated ketone led to 177, from which the agelastatin core was secured by cyclization upon treatment with caesium carbonate, generating 178. Late-stage pyrrole bromination of 178 with N-bromosuccinimide led to 179, which ultimately afforded (−)-agelastatin A by rhodium-catalysed isomerization of the allyl urea to a vinyl urea and subsequent acid hydrolysis. The thirteen-step synthesis afforded this alkaloid in 4 % overall yield. Additionally, a racemic version of this synthesis was reported affording (±)-agelastatin A in 7 % yield over thirteen steps.40

Details are in the caption following the image

Siopa and Afonso's asymmetric total synthesis of (−)-agelastatin A (2022).40

9 Summary and Outlook

The strategies presented above give an overview of the total synthesis developed in Portugal in the past five decades. Over time, and in parallel with the progression of the field of total synthesis, these strategies have in general evolved to achieve higher levels of complexity. Despite the length of certain syntheses featuring multiple manipulations of protecting groups and functional groups interconversions, they nevertheless represent pioneering efforts in many instances and the first total synthesis of specific natural products. Recent strategies have produced more sophisticated syntheses in which a key methodology is developed, such as the Candeias-Gois’ synthesis of viridicatin and Afonso's syntheses of agelastatin A. Furthermore, the ability of recent strategies to allow for the synthesis of natural product analogues highlights the importance of creativity in total synthesis, providing bioactive substances and promising candidates for drug discovery.

With numerous challenges and a vast array of natural products waiting to be synthesised, the field of total synthesis of natural products presents many opportunities for innovative breakthroughs through the development of new reactions, reagents, and catalysts. We hope to encourage synthetic chemists–especially those in Portugal–to consider this field and contribute to the advancement and implementation of creative synthetic strategies.

Acknowledgments

We thank financial support by Fundação para a Ciência e a Tecnologia (FCT) through projects UIDB/00100/2020, UIDP/00100/2020, LA/P/0056/2020 and PTDC/QUI-QOR/1786/2021. J.A.S.C. thanks FCT for Scientific Employment Stimulus 2020/02383/CEECIND.

    Conflict of interest

    The authors declare no conflict of interest.

    Data Availability Statement

    Data sharing is not applicable to this article as no new data were created or analyzed in this study.

    Biographical Information

    Duarte B. Clemente received his B.Sc. in Chemistry in 2020 and M.Sc. in Chemistry in 2022 from the Faculty of Sciences, University of Lisbon, where he is currently a PhD Student in organic chemistry, under the supervision of Dr. Jaime A. S. Coelho.

    Biographical Information

    Jaime A. S. Coelho obtained his B.Sc. (2008) and M.Sc. degrees in Chemistry (2010) from Instituto Superior Técnico, University of Lisbon. He then completed his Ph.D. in 2014 at Faculty of Pharmacy, University of Lisbon under the guidance of Prof. Carlos A. M. Afonso, and held a Visiting Student appointment at Max-Planck-Institut für Kohlenforschung in the laboratory of Prof. Nuno Maulide. After a postdoctoral experience with Professor F. Dean Toste at University of California, Berkeley he joined Faculty of Sciences, University of Lisbon as an Assistant Researcher. His current research interests include the development of new methodologies for the synthesis of biologically active compounds.