Fruitful Decades for Canthin-6-ones from 1952 to 2015: Biosynthesis, Chemistry, and Biological Activities

In this review, more than 60 natural canthin-6-one alkaloids and their structures are considered. The biosynthesis, efficient and classic synthetic approaches, and biological activities of canthin-6-one alkaloids, from 1952 to 2015, are discussed. From an analysis of their structural properties and an investigation of the literature, possible future trends for canthin-6-one alkaloids are proposed. The information reported will be helpful in future research on canthin-6-one alkaloids.

This review provides a broad overview of canthin-6-ones. The first section describes the biosynthetic line and the relevant biosynthetic assembly lines of canthin-6-one alkaloids, the second section considers classic and efficient synthetic methods, and the last section summarizes the excellent biological activity of canthin-6-one alkaloids.

Scheme 1.
Biosynthetic mechanism leading to canthin-6-one alkaloids, and the supporting feeding experiments.

Biosynthetic Assembly Line
Based on the intimate biochemical mechanisms of the biosynthetic pathways, the first nanoparticle system to mimic the relevant biosynthetic assembly lines to canthin-6-one was elucidated by Cebrián-Torrejón et al. in 2013 (Scheme 2) [26]. The strategy used relied on: (i) the covalent linkage Scheme 1. Biosynthetic mechanism leading to canthin-6-one alkaloids, and the supporting feeding experiments.

Biosynthetic Assembly Line
Based on the intimate biochemical mechanisms of the biosynthetic pathways, the first nanoparticle system to mimic the relevant biosynthetic assembly lines to canthin-6-one was elucidated by Cebrián-Torrejón et al. in 2013 (Scheme 2) [26]. The strategy used relied on: (i) the covalent linkage of infractine (7) by "click" chemistry to poly(ethylene glycol); (ii) the use of 7-PEG-OH as a macroinitiator for the ring-opening polymerization of lactide to form 7-PEG-b-PLA copolymer; (iii) the formation of 7-PEG-b-PLA nanoparticles from the self-assembly of 7-PEG-b-PLA in aqueous solution; and (iv) the potential biomimetic release of canthin-6-one (1) from the nanoparticles.

Chemistry
The canthin-6-one alkaloids have been synthesized via different approaches by many researchers [2,. In order to use these methods to guide our review more clearly, we categorized classic and efficient synthetic methods according to their key reaction steps.

Bischer-Napieralski Reaction
The use of the Bischer-Napieralski reaction to synthesize canthin-6-one alkaloids was first reported in 1966 [27], and Soriano-Agaton et al. reported a more recent synthesis in 2005 (Scheme 3) [2]. The overall yield of canthin-6-one in the later synthesis was increased to 76.83% in four steps by Soriano-Agaton.

Chemistry
The canthin-6-one alkaloids have been synthesized via different approaches by many researchers [2,. In order to use these methods to guide our review more clearly, we categorized classic and efficient synthetic methods according to their key reaction steps.

Bischer-Napieralski Reaction
The use of the Bischer-Napieralski reaction to synthesize canthin-6-one alkaloids was first reported in 1966 [27], and Soriano-Agaton et al. reported a more recent synthesis in 2005 (Scheme 3) [2]. The overall yield of canthin-6-one in the later synthesis was increased to 76.83% in four steps by Soriano-Agaton.

Pictet-Spengler Reaction
The Pictet-Spengler reaction was first used to synthesize canthin-6-one alkaloids by Mitscher et al. in 1975 [73], and many syntheses of canthin-6-one that employ this reaction have been reported since then. Czerwinski et al. [54] reported an efficient synthetic approach to canthin-6-one via the Pictet-Spengler reaction in 2003 (Scheme 4), obtaining an overall yield of 46.74%.

Diels-Alder Reaction
In 1992, Snyder and co-workers reported an elegant strategy for accessing the canthine skeleton through using indole as a dienophile in an intramolecular inverse electron demand Diels-Alder (IEDDA) reaction [68]. Subsequently, in 2003, Lindsley et al. [70] reported a "one-pot" microwave-mediated synthesis of canthin-6-one analogs via the IEDDA reaction (Scheme 5), in which the overall yield was 48% in two steps.

Pictet-Spengler Reaction
The Pictet-Spengler reaction was first used to synthesize canthin-6-one alkaloids by Mitscher et al. in 1975 [73], and many syntheses of canthin-6-one that employ this reaction have been reported since then. Czerwinski et al. [54] reported an efficient synthetic approach to canthin-6-one via the Pictet-Spengler reaction in 2003 (Scheme 4), obtaining an overall yield of 46.74%.

Pictet-Spengler Reaction
The Pictet-Spengler reaction was first used to synthesize canthin-6-one alkaloids by Mitscher et al. in 1975 [73], and many syntheses of canthin-6-one that employ this reaction have been reported since then. Czerwinski et al. [54] reported an efficient synthetic approach to canthin-6-one via the Pictet-Spengler reaction in 2003 (Scheme 4), obtaining an overall yield of 46.74%.

Diels-Alder Reaction
In 1992, Snyder and co-workers reported an elegant strategy for accessing the canthine skeleton through using indole as a dienophile in an intramolecular inverse electron demand Diels-Alder (IEDDA) reaction [68]. Subsequently, in 2003, Lindsley et al. [70] reported a "one-pot" microwave-mediated synthesis of canthin-6-one analogs via the IEDDA reaction (Scheme 5), in which the overall yield was 48% in two steps.

Diels-Alder Reaction
In 1992, Snyder and co-workers reported an elegant strategy for accessing the canthine skeleton through using indole as a dienophile in an intramolecular inverse electron demand Diels-Alder (IEDDA) reaction [68]. Subsequently, in 2003, Lindsley et al. [70] reported a "one-pot" microwave-mediated synthesis of canthin-6-one analogs via the IEDDA reaction (Scheme 5), in which the overall yield was 48% in two steps.

Pictet-Spengler Reaction
The Pictet-Spengler reaction was first used to synthesize canthin-6-one alkaloids by Mitscher et al. in 1975 [73], and many syntheses of canthin-6-one that employ this reaction have been reported since then. Czerwinski et al. [54] reported an efficient synthetic approach to canthin-6-one via the Pictet-Spengler reaction in 2003 (Scheme 4), obtaining an overall yield of 46.74%.

Diels-Alder Reaction
In 1992, Snyder and co-workers reported an elegant strategy for accessing the canthine skeleton through using indole as a dienophile in an intramolecular inverse electron demand Diels-Alder (IEDDA) reaction [68]. Subsequently, in 2003, Lindsley et al. [70] reported a "one-pot" microwave-mediated synthesis of canthin-6-one analogs via the IEDDA reaction (Scheme 5), in which the overall yield was 48% in two steps.

Perkin Reaction
The use of the Perkin reaction to synthesize canthin-6-one alkaloid analogs was reported by Giudice et al. in 1990 (Scheme 7) [75] and, more recently, by Brahmbhatt et al. in 2010 [42]. Despite the relatively low overall yield (48.96% for 29 and 46.24% for 30), the synthesis strategy is relatively simple and low-cost.

Non-Classic Strategy
In 2010, Gollner et al. reported a "non-classic" strategy that focused on the construction of the central B ring (Scheme 8) [61]. The strategy relies on a palladium-catalyzed Suzuki-Miyaura C-C coupling followed by a copper-catalyzed C-N coupling that can be achieved either stepwise or in a new one-pot protocol starting from the appropriate 8-bromo-1,5-naphthyridine. Canthin-6-one (1) and nine analogues were prepared rapidly and in high yields (71%-95%). Ethyl canthin-6-one-1carboxylate has also been efficiently synthesized, by Ioannidou et al. in 2011, from readily prepared ethyl 4-bromo-6-methoxy-1,5-naphthyridine-3-carboxylate in a three-step "non-classic" reaction that focuses on the construction of the central pyrrole (B ring) via a palladium-catalyzed Suzuki-Miyaura coupling followed by a copper-catalyzed C-N coupling; the overall yield was 85% [62].

Perkin Reaction
The use of the Perkin reaction to synthesize canthin-6-one alkaloid analogs was reported by Giudice et al. in 1990 (Scheme 7) [75] and, more recently, by Brahmbhatt et al. in 2010 [42]. Despite the relatively low overall yield (48.96% for 29 and 46.24% for 30), the synthesis strategy is relatively simple and low-cost.

Perkin Reaction
The use of the Perkin reaction to synthesize canthin-6-one alkaloid analogs was reported by Giudice et al. in 1990 (Scheme 7) [75] and, more recently, by Brahmbhatt et al. in 2010 [42]. Despite the relatively low overall yield (48.96% for 29 and 46.24% for 30), the synthesis strategy is relatively simple and low-cost.

Non-Classic Strategy
In 2010, Gollner et al. reported a "non-classic" strategy that focused on the construction of the central B ring (Scheme 8) [61]. The strategy relies on a palladium-catalyzed Suzuki-Miyaura C-C coupling followed by a copper-catalyzed C-N coupling that can be achieved either stepwise or in a new one-pot protocol starting from the appropriate 8-bromo-1,5-naphthyridine. Canthin-6-one (1) and nine analogues were prepared rapidly and in high yields (71%-95%). Ethyl canthin-6-one-1carboxylate has also been efficiently synthesized, by Ioannidou et al. in 2011, from readily prepared ethyl 4-bromo-6-methoxy-1,5-naphthyridine-3-carboxylate in a three-step "non-classic" reaction that focuses on the construction of the central pyrrole (B ring) via a palladium-catalyzed Suzuki-Miyaura coupling followed by a copper-catalyzed C-N coupling; the overall yield was 85% [62].

Non-Classic Strategy
In 2010, Gollner et al. reported a "non-classic" strategy that focused on the construction of the central B ring (Scheme 8) [61]. The strategy relies on a palladium-catalyzed Suzuki-Miyaura C-C coupling followed by a copper-catalyzed C-N coupling that can be achieved either stepwise or in a new one-pot protocol starting from the appropriate 8-bromo-1,5-naphthyridine. Canthin-6-one (1) and nine analogues were prepared rapidly and in high yields (71%-95%). Ethyl canthin-6-one-1-carboxylate has also been efficiently synthesized, by Ioannidou et al. in 2011, from readily prepared ethyl 4-bromo-6-methoxy-1,5-naphthyridine-3-carboxylate in a three-step "non-classic" reaction that focuses on the construction of the central pyrrole (B ring) via a palladium-catalyzed Suzuki-Miyaura coupling followed by a copper-catalyzed C-N coupling; the overall yield was 85% [62].

Perkin Reaction
The use of the Perkin reaction to synthesize canthin-6-one alkaloid analogs was reported by Giudice et al. in 1990 (Scheme 7) [75] and, more recently, by Brahmbhatt et al. in 2010 [42]. Despite the relatively low overall yield (48.96% for 29 and 46.24% for 30), the synthesis strategy is relatively simple and low-cost.

Non-Classic Strategy
In 2010, Gollner et al. reported a "non-classic" strategy that focused on the construction of the central B ring (Scheme 8) [61]. The strategy relies on a palladium-catalyzed Suzuki-Miyaura C-C coupling followed by a copper-catalyzed C-N coupling that can be achieved either stepwise or in a new one-pot protocol starting from the appropriate 8-bromo-1,5-naphthyridine. Canthin-6-one (1) and nine analogues were prepared rapidly and in high yields (71%-95%). Ethyl canthin-6-one-1carboxylate has also been efficiently synthesized, by Ioannidou et al. in 2011, from readily prepared ethyl 4-bromo-6-methoxy-1,5-naphthyridine-3-carboxylate in a three-step "non-classic" reaction that focuses on the construction of the central pyrrole (B ring) via a palladium-catalyzed Suzuki-Miyaura coupling followed by a copper-catalyzed C-N coupling; the overall yield was 85% [62].

Antibacterial
In 2007, O'Donnell et al. reported the in vitro antibacterial activity of canthin-6-one alkaloids, and found minimum inhibitory concentrations (MICs) in the range of 8-82 µg/mL against a panel of fast-growing Mycobacterium species and 8-64 µg/mL against multidrug-resistant and methicillin-resistant strains of Staphylococcus aureus [33]. In the same year, Ostrov et al. [34] reported the use of structure-based molecular docking to identify novel drug-like small molecules. They found that canthin-6-one could dock to two sites of Escherichia coli DNA gyrase, targeting and inhibiting the DNA supercoiling activity of purified E. coli DNA gyrase, although it could not effectively accumulate inside E. coli. Our group has designed and synthesized a series of 3-N-alkylated and 3-N-benzylated canthin-6-ones, and evaluated their in vitro antibacterial activities. Of these compounds, eleven 3-N-substituted canthin-6-ones were found to be the most potent, with MIC values lower than 1.95 (µg/mL) against S. aureus [77].

Antitumor
The significant antitumor activity of canthin-6-one alkaloids has been reported by many researchers. For example, in 2014, Devkota et al. [11] reported a study in which canthin-6-one alkaloids were tested in an Nf1and p53-defective mouse malignant glioma tumor cell line engineered to express a dual reporter ( Table 2). The results indicated that the majority of the canthin-6-one alkaloids inhibited cell growth and exhibited some toxicity. The antitumor mechanism was reported by Dejos et al. in the same year [45]. They found that the primary effect of canthin-6-one is as an antiproliferative, possibly by interfering with the G 2 /M transition. In 2015, Cebrian-Torrejon et al. [78] presented an approach for studying the performance of novel targets able to overcome cancer stem cell chemoresistance. The approach was based on voltammetric data for microparticulate films of natural or synthetic alkaloids from the canthin-6-one series. Table 2. Cell viability and inhibition of proliferation (%) in an Nf1and p53-defective mouse Central Nervous System (CNS) tumor cell line by canthines at 2.0 mg/mL.

Antifungal
Canthin-6-one (1) was reported by Thouvenel et al. in 2003 to exhibit a broad spectrum of activity against Aspergillus fumigatus, A. niger, A. terreus, Candida albicans, C. tropicalis, C. glabrata, C. neoformans, Geotrichum candidum, Saccharomyces cerevisiae, Trichosporon beigelii, T. cutaneum, and T. mentagrophytes var. interdigitale, with MICs between 5.3 and 46 µM [36]. Moreover, in 2005, Soriano-Agaton et al. [2] described the structure -activity relationships for the antifungal activity of canthin-6-one ( Figure 2). The mechanism of action of the antifungal canthin-6-one series was investigated in Saccharomyces cerevisiae by Loiseau et al. in 2008 [79]. In 2013, the antifungal mechanism of canthin-6-one was also reported on by Dejos et al. [38]. Although no novel clues to the mechanism were found, they demonstrated that the major-facilitator-superfamily (MFS)-type transporter Flr1 may be able to reduce sensitivity to canthin-6-one when it is overproduced via a higher gene dosage, and that Flr1-mediated tolerance to canthin-6-one is strictly dependent on the transcription factor Yap1. As Dejos et al. stated, "Although the Yap1-Flr1 pair is not naturally involved in yeast tolerance to canthin-6-one, this study demonstrates that their overexpression can lead to resistance to the chemical stress generated by this drug." T. mentagrophytes var. interdigitale, with MICs between 5.3 and 46 µM [36]. Moreover, in 2005, Soriano-Agaton et al. [2] described the structure -activity relationships for the antifungal activity of canthin-6-one ( Figure 2). The mechanism of action of the antifungal canthin-6-one series was investigated in Saccharomyces cerevisiae by Loiseau et al. in 2008 [79]. In 2013, the antifungal mechanism of canthin-6-one was also reported on by Dejos et al. [38]. Although no novel clues to the mechanism were found, they demonstrated that the major-facilitator-superfamily (MFS)-type transporter Flr1 may be able to reduce sensitivity to canthin-6-one when it is overproduced via a higher gene dosage, and that Flr1-mediated tolerance to canthin-6-one is strictly dependent on the transcription factor Yap1. As Dejos et al. stated, "Although the Yap1-Flr1 pair is not naturally involved in yeast tolerance to canthin-6-one, this study demonstrates that their overexpression can lead to resistance to the chemical stress generated by this drug."

Anti-Inflammatory
The transcription factor NF-κB is a key regulator of many proinflammatory pathways, and therefore its inhibition results in anti-inflammatory effects. Canthin-6-one alkaloids were first found to be NF-κB inhibitors by Tran et al. in 2014 [10]. The IC50 values of 9-hydroxycanthin-6-one and 9-methoxycanthin-6-one were 3.8 and 7.4 µM, respectively. However, the IC50 value of the standard drug parthenolide was only 1.5 µM.

Wnt Signaling Inhibitors
Numerous diseases have been attributed to the aberrant transduction of Wnt signaling, which regulates various processes such as cell proliferation and differentiation, and embryo development. In 2015, 9-hydroxycanthin-6-one was screened for its activity in targeting TCF/β-catenin transcriptional modulating activity with a cell-based luciferase assay by Ohishi et al. [80]. The degradation of β-catenin by 9-hydroxycanthin-6-one was suppressed by GSK3β-siRNA, while 9-hydroxycanthin-6-one decreased β-catenin even in the presence of CK1α-siRNA. These results suggest that 9-hydroxycanthin-6-one inhibits Wnt signaling through the activation of GSK3β, independently of CK1α.

Protein Tyrosine Phosphatase 1B Inhibitors
As a potential therapy for diabetes, protein tyrosine phosphatase 1B (PTP1B) inhibitors have attracted considerable attention. In 2015, Sasaki et al. [81] reported that compound 40 (Figure 3) is the competitive PTP1B inhibitor, with the best inhibitory selectivity of PTP1B and other protein tyrosine phosphatase (PTPs), and showed in cell-based assays that it promotes activity in the insulin signaling pathway.

Anti-Inflammatory
The transcription factor NF-κB is a key regulator of many proinflammatory pathways, and therefore its inhibition results in anti-inflammatory effects. Canthin-6-one alkaloids were first found to be NF-κB inhibitors by Tran et al. in 2014 [10]. The IC 50 values of 9-hydroxycanthin-6-one and 9-methoxycanthin-6-one were 3.8 and 7.4 µM, respectively. However, the IC 50 value of the standard drug parthenolide was only 1.5 µM.

Wnt Signaling Inhibitors
Numerous diseases have been attributed to the aberrant transduction of Wnt signaling, which regulates various processes such as cell proliferation and differentiation, and embryo development. In 2015, 9-hydroxycanthin-6-one was screened for its activity in targeting TCF/β-catenin transcriptional modulating activity with a cell-based luciferase assay by Ohishi et al. [80]. The degradation of β-catenin by 9-hydroxycanthin-6-one was suppressed by GSK3β-siRNA, while 9-hydroxycanthin-6-one decreased β-catenin even in the presence of CK1α-siRNA. These results suggest that 9-hydroxycanthin-6-one inhibits Wnt signaling through the activation of GSK3β, independently of CK1α.

Protein Tyrosine Phosphatase 1B Inhibitors
As a potential therapy for diabetes, protein tyrosine phosphatase 1B (PTP1B) inhibitors have attracted considerable attention. In 2015, Sasaki et al. [81] reported that compound 40 (Figure 3) is the competitive PTP1B inhibitor, with the best inhibitory selectivity of PTP1B and other protein tyrosine phosphatase (PTPs), and showed in cell-based assays that it promotes activity in the insulin signaling pathway. T. mentagrophytes var. interdigitale, with MICs between 5.3 and 46 µM [36]. Moreover, in 2005, Soriano-Agaton et al. [2] described the structure -activity relationships for the antifungal activity of canthin-6-one ( Figure 2). The mechanism of action of the antifungal canthin-6-one series was investigated in Saccharomyces cerevisiae by Loiseau et al. in 2008 [79]. In 2013, the antifungal mechanism of canthin-6-one was also reported on by Dejos et al. [38]. Although no novel clues to the mechanism were found, they demonstrated that the major-facilitator-superfamily (MFS)-type transporter Flr1 may be able to reduce sensitivity to canthin-6-one when it is overproduced via a higher gene dosage, and that Flr1-mediated tolerance to canthin-6-one is strictly dependent on the transcription factor Yap1. As Dejos et al. stated, "Although the Yap1-Flr1 pair is not naturally involved in yeast tolerance to canthin-6-one, this study demonstrates that their overexpression can lead to resistance to the chemical stress generated by this drug."

Anti-Inflammatory
The transcription factor NF-κB is a key regulator of many proinflammatory pathways, and therefore its inhibition results in anti-inflammatory effects. Canthin-6-one alkaloids were first found to be NF-κB inhibitors by Tran et al. in 2014 [10]. The IC50 values of 9-hydroxycanthin-6-one and 9-methoxycanthin-6-one were 3.8 and 7.4 µM, respectively. However, the IC50 value of the standard drug parthenolide was only 1.5 µM.

Wnt Signaling Inhibitors
Numerous diseases have been attributed to the aberrant transduction of Wnt signaling, which regulates various processes such as cell proliferation and differentiation, and embryo development. In 2015, 9-hydroxycanthin-6-one was screened for its activity in targeting TCF/β-catenin transcriptional modulating activity with a cell-based luciferase assay by Ohishi et al. [80]. The degradation of β-catenin by 9-hydroxycanthin-6-one was suppressed by GSK3β-siRNA, while 9-hydroxycanthin-6-one decreased β-catenin even in the presence of CK1α-siRNA. These results suggest that 9-hydroxycanthin-6-one inhibits Wnt signaling through the activation of GSK3β, independently of CK1α.

Protein Tyrosine Phosphatase 1B Inhibitors
As a potential therapy for diabetes, protein tyrosine phosphatase 1B (PTP1B) inhibitors have attracted considerable attention. In 2015, Sasaki et al. [81] reported that compound 40 (Figure 3) is the competitive PTP1B inhibitor, with the best inhibitory selectivity of PTP1B and other protein tyrosine phosphatase (PTPs), and showed in cell-based assays that it promotes activity in the insulin signaling pathway.

Conclusions and Future Prospects
The main achievements in the study of canthin-6-one alkaloids over the period of 1952 to 2015 have been reviewed, with emphasis on the biosynthesis, chemistry, and biological activities of these compounds. The low toxicity and good biological activities of canthin-6-one alkaloids mean that there is potential for them to be developed into new drugs. New research by Doménech-Carbó et al. has shown that microparticulate films of canthin-6-one on glassy carbon electrodes could yield separate voltammetric signals for dsDNA, ssDNA, and G-quadruplex DNA, different degrees of DNA methylation, and biomimetic nucleosomal DNA, with a detection limit of 10´5 M [49]. Moreover, excellent photophysical properties of canthin-6-one have been recently reported by Taniguchi et al. [15].
Further research by international groups is required to: (i) develop efficient and environmentally sustainable synthetic strategies for producing canthin-6-one alkaloids on a large scale; (b) elucidate the structure-activity relationships and mechanisms of the different biological activities of these compounds; (c) explore their biological activities; and (d) identify electrochemical and photochemical applications.