Bisindole Alkaloids from the Alstonia Species: Recent Isolation, Bioactivity, Biosynthesis, and Synthesis

Bisindoles are structurally complex dimers and are intriguing targets for partial and total synthesis. They exhibit stronger biological activity than their corresponding monomeric units. Alkaloids, including those containing C-19 methyl-substitution in their monomeric units, their synthetic derivatives, and their mismatched pairs can be attractive targets for synthesis and may unlock better drug targets. We herein discuss the isolation of bisindoles from various Alstonia species, their bioactivity, putative biosynthesis, and synthesis. The total synthesis of macralstonidine, macralstonine, O-acetylmacralstonine, and dispegatrine, as well as the partial synthesis of alstonisidine, villalstonine, and macrocarpamine are also discussed in this review. The completion of the total synthesis of pleiocarpamine by Sato et al. completes the formal synthesis of the latter two bisindoles.


Introduction
Nature has been a substantial and sustainable pool of biologically active compounds. Since ancient times natural product extracts (in crude form) have been used in traditional and folk medicines in many countries. In modern times pure (isolated) natural products and their derivatives play an important role in drug discovery, as indicated by their prevalence in approved drugs for clinical use. Out of the 1881 newly FDA-approved drugs over the last four decades (1 January 1981 to 30 September 2019), a significant portion comprising 506 (26.9%) were either natural products or derived from or inspired by natural products [1]. It is expected that the advent of modern and innovative technologies such as computational software, cheminformatics, artificial intelligence, automation, and quantum computing will further boost natural product-based drug discovery. A synergy among these technological milestones would accelerate hit to lead to clinic pathways of drug discovery, and natural products are expected to remain an important source [2]. Moreover, pharmacophores and their unique stereochemical interactions with natural products may stimulate more demanding targets such as protein-protein interactions in the near future and open up a new avenue in modern drug discovery [3]. The majority of biologically active natural products are produced in plants, known traditionally as medicinal plants.
Alstonia, a major genus in the Apocynaceae family of plants, has nearly 155 species and is found all over the world [4]. Robert Brown named it in 1811 in honor of Charles Alston (1685-1760), an eminent botanist at the University of Edinburgh [4]. The Alstonia genus' trees and shrubs are prevalent in the tropical and subtropical parts of Africa, Asia, and Australia [5]. They contribute significant pharmacological activity, including anticancer, antileishmanial, antimalarial, antitussive, antiviral, antiarthritic, and antibacterial Table 1. Indole alkaloids (monomeric units) in bisindoles from Alstonia species that are reviewed herein.

Types of Monomeric Units Present in Bisindoles
Alstonia Species

Plausible Biogenetic Pathway to (-)-Lumusidines A-D (14-17)
The conjugate addition of the C-10 carbon atom of alstophylline 28 by a Michael reaction onto the α, β unsaturated aldehyde 59 (E-ring opened talcarpine derivative) could potentially give hydroxy ketone bisindole alkaloid 61 ( Figure 3). Alcohol 61 on subsequent closing of the E-ring to form a hemiacetal gives (-)-lumusidine B 15, which after dehydration gives (-)-lumusidine A 14. Similarly, the C-12 carbon atom of alstophylline monomer 28 could add to the α, β unsaturated macroline counterpart 5 to furnish a hydroxy ketone (not shown). The cyclization of the hydroxy ketone in a similar way to give a hemiacetal 62 would follow. Subsequently, the dehydration of hemiacetal 62 would likely form lumusidine D 17. The lumusidine C 16 is assumed to be an artifact formed from hemiacetal 63 (a closed form of (-)-perhentinine 39) as ethanol was used during the extraction [18]. However, again this biosynthetic pathway could occur just as likely via a Friedel-Crafts alkylation of the oxonium ion of 59 (cyclization followed by loss of water) to the olefin of precursor 28 (see Section 4.4 with representative examples of (+)-macralstonine 24 and (+)-lumutinine A 18) [6].

Possible Alternative Mechanism of Bisindole Formation of (+)-Lumutinine A 16 and (+)-Macralstonine 24 as Representative Examples via a Friedel-Crafts Alkylation Process as Suggested by Fukuyama
In addition to the Michael addition process described above, there is another potential mechanism for the coupling of ring-A oxygenated macroline-type alkaloids with macroline 5 that involves a Friedel-Crafts alkylation process stabilized by an oxonium ion [6]. As shown in the figure below ( Figure 5), an acid-catalyzed intramolecular cyclization of macroline 5 by nucleophilic attack of the C-19 hydroxyl function onto the carbonyl carbon atom in 1, 2-addition fashion would furnish the cyclic hemiacetal 70. The loss of a water molecule would generate an oxonium ion species (71a or 71b). At this stage, an electrophilic attack of the electron-rich aromatic ring of 11-methoxy macroline 67 via the C-10 carbon atom onto the olefin in 71a would generate a cyclic enol ether 72. This cyclic hemiacetal could undergo isomerization, generating an oxonium ion intermediate 73 only to be attacked by water to generate the cyclic ether (as in macralstonine 24). Alternatively, the phenolic hydroxy group at the C-10 carbon atom in 10-hydroxy macroline (see 74) could attack the oxonium ion to form a cyclic ketal (as in lumutinine A 16). This type of mechanism was suggested by Fukuyama for some bisindoles isolated from Alstonia species [6].

Plausible Biogenetic Pathway to (-)-Lumusidines A-D (14-17)
The conjugate addition of the C-10′ carbon atom of alstophylline 28 by a Michael reaction onto the α, β unsaturated aldehyde 59 (E-ring opened talcarpine derivative) could potentially give hydroxy ketone bisindole alkaloid 61 ( Figure 3). Alcohol 61 on subsequent closing of the E-ring to form a hemiacetal gives (-)-lumusidine B 15, which after dehydration gives (-)-lumusidine A 14. Similarly, the C-12′ carbon atom of alstophylline monomer 28 could add to the α, β unsaturated macroline counterpart 5 to furnish a hydroxy ketone (not shown). The cyclization of the hydroxy ketone in a similar way to give a hemiacetal 62 would follow. Subsequently, the dehydration of hemiacetal 62 would likely form lumusidine D 17. The lumusidine C 16 is assumed to be an artifact formed from hemiacetal 63 (a closed form of (-)-perhentinine 39) as ethanol was used during the extraction [18]. However, again this biosynthetic pathway could occur just as likely via a Friedel-Crafts alkylation of the oxonium ion of 59 (cyclization followed by loss of water) to the olefin of precursor 28 (see Section 4.4 with representative examples of (+)-macralstonine 24 and (+)lumutinine A 18) [6].

Possible Alternative Mechanism of Bisindole Formation of (+)-Lumutinine A 16 and (+)-Macralstonine 24 as Representative Examples via a Friedel-Crafts Alkylation Process as Suggested by Fukuyama
In addition to the Michael addition process described above, there is another potential mechanism for the coupling of ring-A oxygenated macroline-type alkaloids with macroline 5 that involves a Friedel-Crafts alkylation process stabilized by an oxonium ion [6]. As shown in the figure below ( Figure 5), an acid-catalyzed intramolecular cyclization of macroline 5 by nucleophilic attack of the C-19 hydroxyl function onto the carbonyl carbon atom in 1, 2-addition fashion would furnish the cyclic hemiacetal 70. The loss of a water molecule would generate an oxonium ion species (71a or 71b). At this stage, an electrophilic attack of the electron-rich aromatic ring of 11-methoxy macroline 67 via the C-10 carbon atom onto the olefin in 71a would generate a cyclic enol ether 72. This cyclic hemiacetal could undergo isomerization, generating an oxonium ion intermediate 73 only to be attacked by water to generate the cyclic ether (as in macralstonine 24). Alternatively, the phenolic hydroxy group at the C-10 carbon atom in 10-hydroxy macroline (see 74) could attack the oxonium ion to form a cyclic ketal (as in lumutinine A 16). This type of mechanism was suggested by Fukuyama for some bisindoles isolated from Alstonia species [6].

Synthesis of Representative Monomeric Units of Bisindole Alkaloids from Alstonia
Species Discussed Herein, Required for Bisindole Synthesis 5.1. The Total Synthesis of (±)-Pleiocarpamine 32 (+)-Pleiocarpamine 32 has been isolated from various species of Alstonia [6,32,36,[48][49][50][51]. The unprecedented (+)-bipleiophylline (not shown) was found to contain two pleiocarpamine 32 units connected by utilizing an aromatic spacer (pyrocatechuic acid). The homo-dimer (+) bipleiophylline was potentially active against drug-sensitive, vincristineresistant human KB and Jurkat cells, while its monomeric precursor (+)-pleiocarpamine 32 was not very active [52]. These findings again demonstrate the increased activity of dimeric alkaloids as compared to their monomeric counterparts. The (+)-pleiocarpamine 32 was thought to be formed biosynthetically by a ring-closing reaction between N1 and C16 of geissoschizine 76 [17]. Recently Sato et al. reported the total synthesis of the C-mavacurine type indole alkaloid, (±)-pleocarpamine 32 via a biomimetic synthesis [17]. A critical step, the direct cyclization between N1 and C16, was thought to be challenging because of the strain generated in the corynanthe pentacyclic framework. This may likely be the reason for the time it took to achieve the total synthesis of pleiocarpamine 32 by many groups since its isolation. In fact, (±)-2,7-dihydropleiocarpamine (not shown) was synthesized in 1993 by Jiménez et al. [53]. However, Sato et al. successfully cyclized the strained structure of 77 by a carbene N-H insertion reaction between the N1 and C16 bonds (Scheme 1) [17].
While exploring a suitable catalyst for the metal carbenoid cyclization reaction, the Rh 2 (cap) 4 and Rh 2 (esp) 2 catalyzed reactions lead to decomposition of substrate 77 [17]. Only Rh 2 (OAc) 4 furnished the distinguishable but unanticipated product (not shown) via a D-ring expansion reaction [17]. At room temperature, no reaction happened while employing the rhodium complex (PPh 3 ) 3 RhCl or the copper catalyst Cu(MeCN).BF 4 [17]. Moreover, the use of the JohnPhosAu(MeCN)SbF 6 catalyst was also unsuccessful because it formed the D-ring expanded product (not shown) with β-hydride elimination (not shown) [17]. Sato et al. predicted that a modification of the lone pair on the N(4) of the compound 77 would form a cis-quinolizidine ring making cyclization more favorable by decreasing the distance between N1 and C16. This was the key to the first total synthesis of (±)-pleiocarpamine 32. Thus, substrate 77 was modified to borane salt 83 by using the BH 3 -THF complex to react at the N b nitrogen atom [17]. Finally, cyclization via a carbene N-H insertion was accomplished by treatment of cyclization precursor 83 with Rh 2 (cap) 4 (20 mol %) at room temperature, and the products so obtained were an inseparable diastereomeric mixture 84 at C 16 (ca,7:1) in 55% yield [17]. To remove the borane protecting group, the diastereomeric mixture 84 was refluxed with trimethylamine N-oxide in methanol, which furnished (±)-pleiocarpamine 32 and epi-pleiocarpamine 85 [17]. While exploring a suitable catalyst for the metal carbenoid cyclization reaction, the Rh2(cap)4 and Rh2(esp)2 catalyzed reactions lead to decomposition of substrate 77 [17]. Only Rh2(OAc)4 furnished the distinguishable but unanticipated product (not shown) via a D-ring expansion reaction [17]. At room temperature, no reaction happened while employing the rhodium complex (PPh3)3RhCl or the copper catalyst Cu(MeCN).BF4 [17]. Moreover, the use of the JohnPhosAu(MeCN)SbF6 catalyst was also unsuccessful because it formed the D-ring expanded product (not shown) with β-hydride elimination (not

Biosynthetic Relations among the Sarpagine/Macroline/Ajmaline Family of Alkaloids
Sarpagine, macroline, and ajmaline indole alkaloids are biosynthetically and structurally related (Figure 7) [13]. We herein follow the biogenetic numbering system of LeMen and Taylor to discuss their potential formation [55]. The sarpagine series consists of attachment of the C-21 carbon atom to the N(4) bond, while the macroline series lack this bond. The stereocenters in the sarpagine alkaloids at C-3 (S), C-5 (R), C-15 (R), and C-16 (R) are identical to the macroline series at the respective carbon atoms. The ajmaline series also contain a C-7 and C-17 carbon atom linkage. In comparison with the ajmaline series, the sarpagine family at C-16 is R but antipodal in the ajmaline series. Both the sarpagine and the ajmaline series consist of the characteristic quinuclidine ring and the C-5 and C-16 carbon atom linkage. Sarpagine and ajmaline alkaloids both contain the characteristic C-21 and N(4) bond [13,56]. Stöckigt et al. reported the conversion of 16-epi-vellosimine 86 (sarpagine) into the vinorine 87 (ajmaline) structure (Figure 7) by employing the acetyl CoA-dependent vinorine synthase enzyme [57][58][59][60]. The macroline and sarpagine alkaloids are biosynthetically related, as shown in Figure 7. Either of the β-keto quaternary ammonium intermediates 88 or 89 can be furnished by 1, 4 Michael addition of the N(4) nitrogen atom of macroline onto the α, β-unsaturated carbonyl system 5 or 90 at C-21. Similarly, α-hydroxy quaternary ammonium compounds 91 and 92 can be formed by direct 1, 2 addition of the N(4) nitrogen atom to the C-19 carbonyl carbon of 5 and 90 [13]. In contrast, the retro-Michael reaction of TBS protected intermediate 89, followed by removal of the silyl protecting group converted sarpagine 89 into the macroline unit 5 in the studies from LeQuesne et al. [61] and Cook et al. [62]. The sarpagine/macroline/ajmaline group of indole alkaloids have been isolated principally from higher plants, including the Apocynaceae family and the genera Alstonia and Rauwolfia [13]. The intriguing molecular complexity, interconversions, and their divergent biological activity prompted the successful syntheses of several of these indole alkaloids [7,63] and bisindole alkaloids [45,64,65] as mentioned earlier.

An Improved Total Synthesis of (-)-Alstophylline 28
Liao et al. [66] improved the synthesis of (-)-Alstophylline 28 which, has been isolated from various Alstonia species including A. angustifolia [36], A. glabriflora Mgf. [31], and A. macrophylla [30,67,68]. Liu et al. accomplished a large-scale enantiospecific total synthesis of intermediates, which later resulted in the total synthesis of (-)-alstophylline 28 [69]. Later, Liao et al. improved the total synthesis of (-)-alstophylline 28 employing modified Wacker reaction conditions, which provided a faster route for synthesis of macroline-related indole alkaloids (Scheme 2) [66]. The 6-methoxy tetracyclic ketone 94 was synthesized employing the method using the published procedures [69][70][71], which later was converted into the N b -BH 3 adduct 95 using the method developed by Liu et al. earlier [69]. The N b -BH 3 adduct 95 so formed was refluxed with sodium bicarbonate in methanol overnight to get alcohol 96 in 92% yield. One equivalent of the 2-iodozybenzoic acid (IBX) converted (at high temperature) the alcohol functional group in 96 into the ketone 97 in 85% yield. The addition of three more equivalents of IBX at reflux permitted the formation of the new alkaloid 6-oxoalstophylline 98. The N b -quaternization of ketone 97 was carried out using MeI/THF, which was followed by a retro-Michael reaction in the presence of K 2 CO 3 /THF to furnish α,β-unsaturated ketone 99 in 90% yield. Finally, the (-)-alstophylline 28 was obtained using the Wacker-related process (Pd II) modified by Tsuji et al. [72] and Cook et al. [66] from ketone 99 in 55% yield. Importantly, the modified Wacker-related process is also an improved route to several indole alkaloids with the enone system including alstonerine 100 [6]. In addition, the key oxidation reaction provides a shorter route from 97 to 6-oxoalstophylline 98.
Molecules 2021, 26, x FOR PEER REVIEW 27 of 49 85% yield. The addition of three more equivalents of IBX at reflux permitted the formation of the new alkaloid 6-oxoalstophylline 98. The Nb-quaternization of ketone 97 was carried out using MeI/THF, which was followed by a retro-Michael reaction in the presence of K2CO3/THF to furnish α,β-unsaturated ketone 99 in 90% yield. Finally, the (-)-alstophylline 28 was obtained using the Wacker-related process (Pd II) modified by Tsuji et al. [72] and Cook et al. [66] from ketone 99 in 55% yield. Importantly, the modified Wacker-related process is also an improved route to several indole alkaloids with the enone system including alstonerine 100 [6]. In addition, the key oxidation reaction provides a shorter route from 97 to 6-oxoalstophylline 98.
Later, Liao et al. improved the total synthesis of (+)-macroline 5 in higher overall yield and fewer steps. The tetracyclic ketone 107, obtained on the large scale by following the procedure developed in Milwaukee, which underwent N b -alkylation with (Z)-1-bromo-2-iodo-2-butene (not shown), was followed by a palladium catalyzed stereospecific αvinylation to provide pentacyclic ketone 108 (Scheme 4) [76]. The ketone 108 underwent the usual one-carbon homologation via a Wittig olefination, and this was followed by acidic hydrolysis to obtain the pentacyclic aldehyde 109. The aldehyde functional group in 109 was reduced using sodium borohydride to provide the first enantiospecific synthesis of (+)-affinisine (not shown) in 90% yield. Then triisopropylsilyl (TIPS) triflate and 2,6lutidine were used to protect the primary alcohol of (+)-affinisine as a TIPS ether 110 in 90% yield. The TIPS ether 110 was employed for the hydroboration-oxidation reaction, which furnished the desired secondary alcohol 111 in 86% yield. The alcohol 111, so formed, was oxidized to a ketone 112 using the Dess-Martin periodinane reagent in 82% yield. The ketone 112 was quaternized at the N b -nitrogen atom by treatment with iodomethane in THF at 0 • C. The stable macroline equivalent 113 was obtained after treatment of N b-methyl iodide salt with potassium tert-butoxide in ethanol, after which it was treated with tetrabutylammonium fluoride to furnish macroline 5 in 86% yield (99% ee) after removal of the TIPS group [76]. which furnished the desired secondary alcohol 111 in 86% yield. The alcohol 111, so formed, was oxidized to a ketone 112 using the Dess-Martin periodinane reagent in 82% yield. The ketone 112 was quaternized at the Nb-nitrogen atom by treatment with iodomethane in THF at 0 °C. The stable macroline equivalent 113 was obtained after treatment of Nb-methyl iodide salt with potassium tert-butoxide in ethanol, after which it was treated with tetrabutylammonium fluoride to furnish macroline 5 in 86% yield (99% ee) after removal of the TIPS group [76]. Scheme 4. Improved synthesis of (+)-macroline 5. Scheme 4. Improved synthesis of (+)-macroline 5.

Stereospecific Access to the Macroline Core 5
Recently, Kadam et al. developed another route to the pentacyclic core of macroline 5-related alkaloids. Starting from commercially available L-tryptophan (not shown), the core architecture of macroline 5 was accessed via an Ireland-Claisen rearrangement (ICR)/Pictet-Spengler cyclization protocol (Scheme 5) [82]. The L-tryptophan derivative 114 was subjected to Arndt-Eistert conditions to implement a one-carbon homologation to furnish acid 115. The acid 115 was subsequently coupled with allylic alcohol 116 under standard amide coupling conditions to furnish the allylic ester 117, the substrate for the Ireland-Claisen rearrangement. The allylic ester 117 produced the corresponding carboxylic acids (not shown) under ICR as an inseparable mixture that was converted into the corresponding esters 118A/A' by treating the ICR mixture. The esters were isolated in a 9:1 ratio of diastereomers that was separable. The ester function in 118A was reduced by DIBAL-H, and the so formed primary alcohol was protected with a TBS group to provide silyl ether 119. The olefin 119 was subjected to a hydroboration-oxidation sequence and this was followed by oxidation of the so formed alcohol with Dess-Martin periodinane to furnish aldehyde 120. Upon desilylation with TBAF, the alcohol reacted intramolecularly with the proximate aldehyde function forming the corresponding hemiacetal (not shown).

Biomimetic Partial Synthesis of (+)-Dispegatrine 131
Although the focus of this review is bisindoles from Alstonia species, dispegatrine 131, a hypotensive bisindole, is comprised of two sarpagine units. It is discussed here because some of the monomers such as sarpagine are commonly found in many Alstonia species. In the biomimetic partial synthesis of (+)-dispegatrine 131 by Yu et al., the monomer spegatrine 132 was subjected to an oxidative phenolic coupling (Scheme 8). Although the yield of this coupling was very poor, it served as proof of the concept that dispegatrine 131 may actually originate biosynthetically via oxidative phenolic coupling, although the chemical yield was 0.25%. This partial synthesis confirmed the structure of (+)-dispegatrine 131 [86]. However, no other diastereomer was observed, and the axial chirality across C9-C9 could not be confirmed. In addition, due to the lack of X-ray crystals, the isolation chemists could not assign the important axial chirality at C9-C9 [86]. Molecules 2021, 26, x FOR PEER REVIEW 33 of 49 Scheme 7. Biomimetic partial synthesis of (-)-alstonisidine 3

Biomimetic Partial Synthesis of (+)-Dispegatrine 131
Although the focus of this review is bisindoles from Alstonia species, dispegatrine 131, a hypotensive bisindole, is comprised of two sarpagine units. It is discussed here because some of the monomers such as sarpagine are commonly found in many Alstonia species. In the biomimetic partial synthesis of (+)-dispegatrine 131 by Yu et al., the monomer spegatrine 132 was subjected to an oxidative phenolic coupling (Scheme 8). Although the yield of this coupling was very poor, it served as proof of the concept that dispegatrine 131 may actually originate biosynthetically via oxidative phenolic coupling, although the chemical yield was 0.25%. This partial synthesis confirmed the structure of (+)-dispegatrine 131 [86]. However, no other diastereomer was observed, and the axial chirality across C9-C9′ could not be confirmed. In addition, due to the lack of X-ray crystals, the isolation chemists could not assign the important axial chirality at C9-C9′ [86]. Scheme 8. Biomimetic partial synthesis of (+)-dispegatrine 131 by oxidative phenolic coupling by Yu et al. [86].
Enantiospecific Total Synthesis of (+)-dispegatrine 131 in >98% ee It was suggested that internal chiral induction was responsible for the atroposelectivity, which was exploited as an advantage in the first enantiospecific and convergent total synthesis of (+)-dispegatrine 131 from spegatrine 132 by Edwankar et al. [87,88]. The total synthesis of dispegatrine 131 required the 5-methoxy tryptophan unit 133, which was accessed via the general strategy for the synthesis of ring-A oxygenated D-tryptophan Scheme 7. Biomimetic partial synthesis of (-)-alstonisidine 3.

Biomimetic Partial Synthesis of (+)-Dispegatrine 131
Although the focus of this review is bisindoles from Alstonia species, dispegatrine 131, a hypotensive bisindole, is comprised of two sarpagine units. It is discussed here because some of the monomers such as sarpagine are commonly found in many Alstonia species. In the biomimetic partial synthesis of (+)-dispegatrine 131 by Yu et al., the monomer spegatrine 132 was subjected to an oxidative phenolic coupling (Scheme 8). Although the yield of this coupling was very poor, it served as proof of the concept that dispegatrine 131 may actually originate biosynthetically via oxidative phenolic coupling, although the chemical yield was 0.25%. This partial synthesis confirmed the structure of (+)-dispegatrine 131 [86]. However, no other diastereomer was observed, and the axial chirality across C9-C9′ could not be confirmed. In addition, due to the lack of X-ray crystals, the isolation chemists could not assign the important axial chirality at C9-C9′ [86]. Scheme 8. Biomimetic partial synthesis of (+)-dispegatrine 131 by oxidative phenolic coupling by Yu et al. [86].
Enantiospecific Total Synthesis of (+)-dispegatrine 131 in >98% ee It was suggested that internal chiral induction was responsible for the atroposelectivity, which was exploited as an advantage in the first enantiospecific and convergent total synthesis of (+)-dispegatrine 131 from spegatrine 132 by Edwankar et al. [87,88]. The total synthesis of dispegatrine 131 required the 5-methoxy tryptophan unit 133, which was accessed via the general strategy for the synthesis of ring-A oxygenated D-tryptophan Scheme 8. Biomimetic partial synthesis of (+)-dispegatrine 131 by oxidative phenolic coupling by Yu et al. [86].
Enantiospecific Total Synthesis of (+)-dispegatrine 131 in >98% ee It was suggested that internal chiral induction was responsible for the atroposelectivity, which was exploited as an advantage in the first enantiospecific and convergent total synthesis of (+)-dispegatrine 131 from spegatrine 132 by Edwankar et al. [87,88]. The total synthesis of dispegatrine 131 required the 5-methoxy tryptophan unit 133, which was accessed via the general strategy for the synthesis of ring-A oxygenated D-tryptophan unit 134 previously developed in Milwaukee (Scheme 9). L-Valine 135 served as the source of chirality in the Schöllkopf chiral auxiliary 136 and was employed in a large-scale reaction (500 g). After obtaining the required chiral starting material, D-tryptophan ethyl ester 133 (which could also be obtained from the Boc protected 5-methoxy indole 137 (Scheme 9)), the large-scale Pictet-Spengler/Dieckmann protocol developed in Milwaukee was employed to gain access to the key tetracyclic ketone 138 in excellent yield and >98% enantiomeric excess. This tetracyclic ketone 138 furnished the sarpagine system 139 after N b -alkylation and α-vinylation, both of which could also be performed on large scales. The pentacyclic ketone 139 containing the sarpagine architecture was further functionalized. A one-carbon homologation via a Wittig olefination, followed by hydrolysis under acidic conditions, furnished 10-methoxy vellosimine 140 for the first time (>98% ee). Reduction of the C-16 formyl function with NaBH 4 furnished (+)-lochnerine 141 in excellent yield. A thallium (III)mediated oxidative coupling facilitated by borane trifluoride diethyl etherate as the Lewis acid provided the dimerized lochnerine 142 as the P(S)-atropodiastereomer. The absolute configuration was confirmed by X-ray analysis with a MoKα source at low temperature. At the conclusion of this total synthesis, the final assault, demethylation of the two methoxy groups followed by quaternization of the N b -nitrogen atoms was performed. Treatment of the lochnerine dimer 142 with boron tribromide etherate furnished the bis-phenolic dimer (not shown) in 80% yield. Quaternization was executed by treating the tertiary amine functions with excess iodomethane, and this was followed by exchange of iodide with chloride as the counter anion with the help of a AgCl additive. This furnished the first and biomimetic total synthesis of the antihypertensive bisindole, the P(S)-atropodiastereomer of (+)-dispegatrine 131 [87,88].

Biomimetic Total Synthesis of (+)-Macralstonidine 23
(+)-Macralstonidine 23 is composed of a macroline 5 and a N a -methyl sarpagine 124 unit. (+)-N a -Methyl sarpagine 124 (sarpagine unit) was isolated from A. spectabilis [31] and its total synthesis was accomplished by Zhao et al., as discussed earlier [83,84]. In a chemical sense, (+)-macralstonidine 23 should arise from the acid-mediated condensation of macroline 5 with (+)-N a -methyl sarpagine 124. When macroline 5 was stirred with (+)-N a -methyl sarpagine 124 under mildly acidic conditions, the so formed cyclic ketal was identical to the authentic sample of macralstonidine 23 [31]. In this case too, the condensation followed by cyclic ketal formation can be explained by two different mechanisms. The straightforward Michael-type cyclization, which was discussed previously, is shown below (Scheme 10). On the other hand, the Friedel-Crafts alkylation-type mechanism that involved stabilization of the macroline enone with an oxonium anion is somewhat more probable from a chemical sense. As shown in Scheme 10, after initial intramolecular reaction of the C-17 hydroxyl function with the macroline ketone 5, followed by acid-catalyzed dehydration, this process produced the oxonium species, which existed in two resonance stabilized forms (71a and 71b). The electron-rich phenolic sarpagine 124 carbon atom could attack this electrophile from the C-9 position, which was ortho to the phenolic group. No alkylation at the C-11 position was observed. The so formed enol ether underwent acid-mediated rearrangement forming an oxonium intermediate 143, which made it susceptible to nucleophilic attack by the phenolic oxygen atom furnishing the cyclic ketal present in macralstonidine 23. Since macroline 5 and N a -methyl sarpagine 124 have both been synthesized in Milwaukee, this constitutes an enantiospecific total synthesis of (+)-macralstonidine 23 [76,79].

Partial Synthesis of (-)-Macrocarpamine 31
Gan et al. accomplished the partial synthesis of (-)-macrocarpamine 31 in stereospecific and enantiospecific fashion via the coupling reaction between the protonated natural (+)pleiocarpamine 32 and synthetic (-)-anhydromacrosalhine-methine 33 (Scheme 12) [44]. The portion-wise addition of six equivalents of (-)-anhydromacrosalhine-methine 33 to (+)-pleiocarpamine 32 in 0.2 N dry HCl/THF, followed by basic workup (10% NH 4 OH), furnished (-)-macrocarpamine 31 in 75% yield [65]. It is important to stir pleiocarpamine 32 in acid to form the iminium ion first and then add (-)-33 to it. Initial treatment of (-)-anhydromacrosalhine-methine 33 and (+)-pleiocarpamine 32 in aqueous and mild acidic conditions (0.2 N HCl) converted enol 33 into its hydrated product, inhibiting the desired product 144 formation. This was reported to be the first coupling reaction between an iminium ion and vinylogous enol ether of such type (33) and the first proton-mediated coupling reaction to obtain a bisindole via the coupling of two monomeric units from the Alstonia genus [65]. The total synthesis of (-)-macrocarpamine 31 has not been reported yet. Since the total synthesis of both of its monomeric units are now successfully completed, the formal total synthesis of (-)-macrocarpamine 31 is completed [17,44].

Improved Partial Synthesis of (+)-Villalstonine 43
Because macroline 5 under acidic conditions forms part of the very stable dihydroalstonerine 105 (Scheme 14), the process was later modified by using the more stable macroline equivalent 146 [45]. The alcohol functional group in the macroline equivalent 5 was protected as its TBDMS ether 146 [79]. The natural (+)-pleiocarpamine 32 was received from Professor LeQuesne and then treated with the more stable macroline equivalent 146 in the presence of 0.2 N aqueous hydrochloric acid and fluoride ion (tert-butylammonium fluoride) for 24 h at 22 °C (Scheme 15). This procedure was followed by a basic workup with aqueous NH4OH to obtain (+)-villalstonine 43 as an oil in 60% condensation yield. Traces of (+)-pleiocarpamine 32 remained on TLC. The desilylation might occur before condensation to give macroline 146, or the condensation may take place first as shown in Scheme 15 [6]. The spectroscopic data, including IR and NMR, TLC, and mass spectrometry of synthetic (+)-villalstonine 43, were identical when compared to the authentic sample separately isolated from A. muelleriana by LeQuesne and Schmid et al. [38,51]. Since the total synthesis of both of its monomeric units are now successfully completed, the formal total synthesis of (+)-villalstonine 43 is completed [47,78].

Improved Partial Synthesis of (+)-Villalstonine 43
Because macroline 5 under acidic conditions forms part of the very stable dihydroalstonerine 105 (Scheme 14), the process was later modified by using the more stable macroline equivalent 146 [45]. The alcohol functional group in the macroline equivalent 5 was protected as its TBDMS ether 146 [79]. The natural (+)-pleiocarpamine 32 was received from Professor LeQuesne and then treated with the more stable macroline equivalent 146 in the presence of 0.2 N aqueous hydrochloric acid and fluoride ion (tert-butylammonium fluoride) for 24 h at 22 °C (Scheme 15). This procedure was followed by a basic workup with aqueous NH4OH to obtain (+)-villalstonine 43 as an oil in 60% condensation yield. Traces of (+)-pleiocarpamine 32 remained on TLC. The desilylation might occur before condensation to give macroline 146, or the condensation may take place first as shown in Scheme 15 [6]. The spectroscopic data, including IR and NMR, TLC, and mass spectrometry of synthetic (+)-villalstonine 43, were identical when compared to the authentic sample separately isolated from A. muelleriana by LeQuesne and Schmid et al. [38,51]. Since the total synthesis of both of its monomeric units are now successfully completed, the formal total synthesis of (+)-villalstonine 43 is completed [47,78]. The biomimetic synthesis of (+)-villalstonine 43 is thought to begin via electrophilic addition of the macroline enone 5 onto pleiocarpamine 32 from the α-face (less hindered) to give the intermediate 145 (Scheme 13) [51]. The intermediate 145 then forms a hemiacetal, after which the newly formed acetal sequentially attacks the iminium ion of the indoline from the less hindered side to furnish the acetal-amino acetal function in (+)-villalstonine 43 [51].

Improved Partial Synthesis of (+)-Villalstonine 43
Because macroline 5 under acidic conditions forms part of the very stable dihydroalstonerine 105 (Scheme 14), the process was later modified by using the more stable macroline equivalent 146 [45]. The alcohol functional group in the macroline equivalent 5 was protected as its TBDMS ether 146 [79]. The natural (+)-pleiocarpamine 32 was received from Professor LeQuesne and then treated with the more stable macroline equivalent 146 in the presence of 0.2 N aqueous hydrochloric acid and fluoride ion (tert-butylammonium fluoride) for 24 h at 22 • C (Scheme 15). This procedure was followed by a basic workup with aqueous NH 4 OH to obtain (+)-villalstonine 43 as an oil in 60% condensation yield. Traces of (+)-pleiocarpamine 32 remained on TLC. The desilylation might occur before condensation to give macroline 146, or the condensation may take place first as shown in Scheme 15 [6]. The spectroscopic data, including IR and NMR, TLC, and mass spectrometry of synthetic (+)-villalstonine 43, were identical when compared to the authentic sample separately isolated from A. muelleriana by LeQuesne and Schmid et al. [38,51]. Since the total synthesis of both of its monomeric units are now successfully completed, the formal total synthesis of (+)-villalstonine 43 is completed [47,78].

C-19 Methyl-Substituted Sarpagine/Macroline/Ajmaline Alkaloids
The C-19 methyl-substituted sarpagine/macroline/ajmaline-type alkaloids are an emerging group of structurally and biosynthetically related alkaloids. To date, more than 70 monomeric indole alkaloids that belong to this sub-family have been isolated. In addition to the monomeric bases, several bisindole alkaloids consisting of at least one monomeric unit with a C-19 methyl group as their obligate constituent have been found in various medicinal plants worldwide (vide infra, Table 2). Bioactive C-19 methyl-substituted sarpagine/macroline bases and related alkaloids occur in various Alstonia species. The reported bioactivity of representative examples is listed herein (vide infra, Figure 8). It can be seen in Figure 8 that these alkaloids contain a stereogenic methyl function at the C-19 position and presumably originate from common or similar biosynthetic precursors. As a result, some common intermediates or core architecture should enable the total synthesis of these alkaloids. As illustrated earlier in Figure 7, macroline, sarpagine, and ajmaline alkaloids are essentially biosynthetic relatives, and they contain the azabicyclo[3.3.1]nonane core 93 present in their structures. Thus, any effective and divergent synthetic strategy should involve robust access to the common intermediates, preferably containing the azabicyclo[3.3.1]nonane core system 93. The total synthesis of bisindoles containing the C-19 methyl function has not previously been reported. On the other hand, there has been notable progress in developing synthetic strategies for practical access to this important class of bioactive indole alkaloids [7,62].

C-19 Methyl-Substituted Sarpagine/Macroline/Ajmaline Alkaloids
The C-19 methyl-substituted sarpagine/macroline/ajmaline-type alkaloids are an emerging group of structurally and biosynthetically related alkaloids. To date, more than 70 monomeric indole alkaloids that belong to this sub-family have been isolated. In addition to the monomeric bases, several bisindole alkaloids consisting of at least one monomeric unit with a C-19 methyl group as their obligate constituent have been found in various medicinal plants worldwide (vide infra, Table 2). Bioactive C-19 methyl-substituted sarpagine/macroline bases and related alkaloids occur in various Alstonia species. The reported bioactivity of representative examples is listed herein (vide infra, Figure 8). It can be seen in Figure 8 that these alkaloids contain a stereogenic methyl function at the C-19 position and presumably originate from common or similar biosynthetic precursors. As a result, some common intermediates or core architecture should enable the total synthesis of these alkaloids. As illustrated earlier in Figure 7, macroline, sarpagine, and ajmaline alkaloids are essentially biosynthetic relatives, and they contain the azabicyclo[3.3.1]nonane core 93 present in their structures. Thus, any effective and divergent synthetic strategy should involve robust access to the common intermediates, preferably containing the azabicyclo[3.3.1]nonane core system 93. The total synthesis of bisindoles containing the C-19 methyl function has not previously been reported. On the other hand, there has been notable progress in developing synthetic strategies for practical access to this important class of bioactive indole alkaloids [7,62].

Enantiospecific Total Synthesis of C-19 Methyl-Substituted Sarpagine/Macroline Alkaloids via the General Strategy Developed in Milwaukee
During the development of an improved and shorter access to the tetracyclic ketone containing the azabicyclo[3.3.1]nonane core architecture, Rahman et al. developed an ambidextrous Pictet-Spengler (P-S)/Dieckmann protocol that provided efficient stereospecific access to the desired common intermediates towards the C-19 methyl-substituted sarpagine/macroline-type monomeric indole alkaloids [92]. The key tetracyclic intermediates for access to both the C-19 (S) or (R)-methyl-substituted alkaloids can now be synthesized either from D-or from L-tryptophan, at will, albeit at improved yield and in a route two steps shorter than the former strategy (Scheme 16). Further studies in this vein resulted in an illustration of the applicability of the so developed ambidextrous P-S strategy by accessing both enantiomers (individually) of the azabicyclo[3.3.1]nonane core structures, while starting from the same tryptophan auxiliary. One of the impor-tant implications of this ambidextrous P-S/Dieckmann protocol is that not only are the natural alkaloids containing both (S) and (R) C-19 methyl-containing alkaloids available but their enantiomers (unnatural alkaloids) can also be accessed using this strategy from the same chiral auxiliary (Scheme 17). To date, this unified general strategy has been illustrated in the total synthesis of more than a dozen alkaloids from this sub-family ( Figure 9) [7,56,62,[93][94][95]. A detailed discussion of the reported total synthesis of the individual monomeric alkaloids from this sub-family is out of the scope of this review. The enantiospecific total synthesis of the first example of sarpagine-type alkaloids isolated from the root-culture of R. serpentina that belong to this group was reported by Edwakar et al. [94]. These two monoterpenoid sarpagine alkaloids, 19(S),20(R)-dihydroperaksine-17-al 147 and 19(S),20(R)-dihydroperaksine 148 ( Figure 9) represented a novel subgroup of sarpagine alkaloids, identified by Stöckigt et al. [96]. Afterwards, several other alkaloids including peraksine 149, macrosalhine-N b -oxide 150, and talcarpine 151 (formal synthesis) from this group were reported by employing the general strategy developed in Milwaukee [93]. Later, as a result of continued studies in this vein, Rahman et al. employed an efficient coppermediated enolate-driven cross-coupling process that replaced the previous palladiumcatalyzed coupling to gain an improved access to the core architecture of this subgroup of alkaloids and illustrated the process in the total synthesis of two monomeric macroline alkaloids, macrocarpines D 152 and E 153 [62]. In 2018, Rahman et al. reported an improved strategy for a two-step shorter and higher overall yield to achieve access to the key tetracyclic ketone intermediates. The total synthesis of a number of sarpagine and macroline alkaloids, including talcarpine 151, N 4 -methyl-N 4 ,21-secotalpinine 154, macrocarpines A-C (155-157), dihydroperaksine 148, and deoxyperaksine 158, was completed and reported in the same paper. This also corrected the reported optical rotations of N 4methyl-N 4 ,21-secotalpinine 154 and macrocarpine A 155 [7]. Afterwards, the focus turned toward the development of the ambidextrous Pictet-Spengler approach for accessing natural alkaloids either from D-or L-tryptophan, at will. Besides, both natural and unnatural enantiomers would be accessible from the same chiral starting material. Several more alkaloids from this subgroup, including the potent NF-kB inhibitor N 4 -methyltalpinine 159, anticancer alkaloids talpinine 160 and O-acetyl talpinine 161, as well as the macrocarpines F-G (162-163), have been synthesized and will be reported in due course (Figure 9) [97]. Scheme 15. Improved partial synthesis of (+)-villalstonine 43

C-19 Methyl-Substituted Sarpagine/Macroline/Ajmaline Alkaloids
The C-19 methyl-substituted sarpagine/macroline/ajmaline-type alkaloids are an emerging group of structurally and biosynthetically related alkaloids. To date, more than 70 monomeric indole alkaloids that belong to this sub-family have been isolated. In addition to the monomeric bases, several bisindole alkaloids consisting of at least one monomeric unit with a C-19 methyl group as their obligate constituent have been found in various medicinal plants worldwide (vide infra, Table 2). Bioactive C-19 methyl-substituted sarpagine/macroline bases and related alkaloids occur in various Alstonia species. The reported bioactivity of representative examples is listed herein (vide infra, Figure 8). It can be seen in Figure 8 that these alkaloids contain a stereogenic methyl function at the C-19 position and presumably originate from common or similar biosynthetic precursors. As a result, some common intermediates or core architecture should enable the total synthesis of these alkaloids. As illustrated earlier in Figure 7, macroline, sarpagine, and ajmaline alkaloids are essentially biosynthetic relatives, and they contain the azabicyclo[3.3.1]nonane core 93 present in their structures. Thus, any effective and divergent synthetic strategy should involve robust access to the common intermediates, preferably containing the azabicyclo[3.3.1]nonane core system 93. The total synthesis of bisindoles containing the C-19 methyl function has not previously been reported. On the other hand, there has been notable progress in developing synthetic strategies for practical access to this important class of bioactive indole alkaloids [7,62].   Any member of this group would be accessible employing this general strategy, albeit most of these alkaloids await their total synthesis. In addition, no bisindole alkaloid from this group has been synthesized, to date, but the building blocks are now in place. The important bioactivity of a number of alkaloids from this group or their unnatural enantiomers (both monomeric and bisindole) by combining different bioactive monomeric units might result in novel chemical entities, which are unavailable in normal biosynthetic processes. In this respect, the enantiomers of the bioactive monomeric C-19 methylated alkaloids would also be available from either of the tryptophan esters at will. Enantiomers may have similar or even better properties depending on their metabolism, and this holds true for bisindoles, as well as their mismatched pairs. Molecules 2021, 26, x FOR PEER REVIEW 44 of 49 Figure 9. Examples of C-19 methylated alkaloids synthesized employing the general strategy developed herein

Unnatural (Enantiomers or Synthetic Derivatives) Alkaloids in Drug Discovery
Natural molecules or compounds are usually chiral molecules that are the secondary metabolites used in plants or organisms, perhaps as defense mechanisms. The plants or organisms have evolved to produce only particular enantiomers, which results in the isolation and screening of only particular enantiomers, i.e., the natural enantiomers. Optically pure drugs from natural sources exhibit desired biological activity usually in particular enantiomeric forms, for example, morphine and reserpine [98]. Anticipated pharmacological effects can be achieved by applying absolute stereochemistry or enantioselectivity of compounds

Unnatural (Enantiomers or Synthetic Derivatives) Alkaloids in Drug Discovery
Natural molecules or compounds are usually chiral molecules that are the secondary metabolites used in plants or organisms, perhaps as defense mechanisms. The plants or organisms have evolved to produce only particular enantiomers, which results in the isolation and screening of only particular enantiomers, i.e., the natural enantiomers. Optically pure drugs from natural sources exhibit desired biological activity usually in particular enantiomeric forms, for example, morphine and reserpine [98]. Anticipated pharmaco-logical effects can be achieved by applying absolute stereochemistry or enantioselectivity of compounds in many enzymatic reactions [98]. Logically, the unnatural enantiomers might be as good as the natural enantiomers or even better in activity and toxicity profile, depending on the metabolism. For example, a drug in WHO's essential list, lamivudine, an unnatural deoxycytidine analog, has saved millions of lives after its approval by the FDA in 1995 [99]. The unnatural L-nucleoside-based drug, emtricitabine, exhibited 100 times greater potency and longer duration of action than its natural D-nucleoside analog [92,99,100]. Among 185 novel small molecules approved by the FDA since 1981 for treatment of cancer, 120 (64.9%) are from a natural source, natural product mimetic, or natural product pharmacophore-inspired synthetic drug [1]. Notably, several C'-20 urea derivates of the robust clinical drug vinblastine are reported to be 10-200 times more active than the natural vinblastine analogs themselves against various human tumor cell lines [101]. When compared to the original compound, unnatural/synthetic drug candidates such as analogs of bryostatin, vincristine, and vinblastine might exhibit excellent biological activity [102][103][104]. The modified vinca alkaloids (fluorine substitution at C10 ) 10 -fluorovinblastine and 10 -fluorovincristine exhibited substantially stronger activity than the natural microtubule inhibitors vinblastine and vincristine against the sensitive and vinblastine-resistant tumor cell lines, in the study from the Boger group [105]. Thus, it is crucial to discover the pharmacological effects of derivatives and unnatural enantiomers of natural products.

Conclusions
Natural products and nature-inspired molecules are expected to continue playing an indispensable role in modern drug discovery. Compared to their monomeric units, bisindoles are more potent in many cases. Bisindole alkaloids including semisynthetic derivatives have been found to possess significant bioactivity, including anticancer, antileishmanial, and antimalarial properties, and thus are promising leads for nature-inspired drug discovery and development. Unnatural medicinal compounds formed by combining bioactive mismatched monomeric units can furnish novel medicinal compounds. Incorporating unnatural enantiomers of monomeric alkaloids into bisindoles can provide access to novel and unnatural bioactive compounds that may have better activity and in vivo stability depending on their metabolism.