MSc Chemistry
Molecular Sciences
Literature Thesis
Reported Synthetic Approaches Toward the
bis-THF Core of Annonaceous Acetogenins
by
Martijn Johannes Patist
UvA: 11369299
VU: 2630045
April 2020
12 EC
May 2019 – August 2019
Supervisor/Examiner:
Examiner:
J. (John) Braun
dr. T. (Tati) Fernández Ibáñez
prof. dr. ir. R.V.A. (Romano) Orru
Synthetic & Bioorganic Chemistry
3
Abstract
Annonaceous acetogenins have cytotoxic effects and a distinct proposed mechanism of action compared to other chemotherapeutical medicines for the treatment of cancer. Here, a comprehensive overview of the synthetic routes toward the bis-THF core in acetogenins are discussed. The bis-THF core building block of several acetogenins are different in multiple aspects. The stereochemistry, substitutions next to the bis-THF unit, and whether the THF units are adjacent or non-adjacent to each other gave a large amount of possibilities for the synthesis toward bis-THF cores in acetogenins. For the adjacent bis-THF building blocks, many strategies are based on a cascade mechanism. However, also non-cascade approaches are discussed. For non-adjacent bis-THF cores, the most approaches required a metathesis of two buildi ng blocks resulting in a bis-THF cored moiety. Although also for this section, multiple synthetic strategies have been discussed.
4
Table of Contents
Introduction ...5
General introduction on Annonaceous Acetogenins ...5
The molecular fingerprint of acetogenins...5
Nomenclature of the bis-THF core ...5
Desymmetrization reactions on C2-symmetrical molecules ...6
Biological activity ...7
Biosynthesis...8
Polyketide pathway in natural products ...8
Post-modification(s) in the biosynthesis toward the THF core ...8
The biosynthesis toward mono-THF cores...9
The biosynthesis toward adjacent bis-THF cores ...9
The biosynthesis toward tris-THF cores and non-adjacent bis-THF cores ... 10
Adjacent bis-THF core ... 11
Cascade methodologies... 11
Double SN2 strategies ... 11
Metal catalyzed strategies ... 17
Iodoetherification ... 24
Epoxide as nucleophile ... 25
Non-Cascade methodologies ... 27
Double SN2 strategies ... 27
Single Substitution strategies ... 30
Non-adjacent bis-THF Core... 32
Without the use of building blocks ... 32
The use of building blocks... 34
Bimolecular diene cross metathesis... 34
Other synthesis from bimolecular building blocks ... 35
Unimolecular silicon tethered ring closing metathesis ... 36
Conclusion ... 40
Recommendations... 41
Acknowledgements ... 41
5
Introduction
General introduction on Annonaceous Acetogenins
Annonaceous acetogenins are natural products found in the Annonaceae plant family and show interesting biological activity against tumor cell lines.1 These secondary metabolites were first isolated by Cole and co-workers, who were able to isolate the acetogenin uvaricin (92, Figure 1) in 1982.2 Since many structurally distinct Annonaceous acetogenins were found, many total syntheses were pursued. In this literature study, a comprehensive overview of the synthetic approaches toward the bis-THF core moieties of the acetogenins known in literature will be discussed.
The molecular fingerprint of acetogenins
Annonaceous acetogenins have a general structure (Figure 1), containing a bis-THF core that is depicted in blue. The core has two THF units attached to each other, however there are also other variations known; the core of some acetogenins consist of one or three attached THF units instead.3,4 The bis-THF-core illustrated in Figure 1, is known as an adjacent bis-THF core moiety. In addition to this bis-THF core, also non-adjacent bis-THF cores are known. Moreover, the R groups could all contain a hydroxyl group. However, R1 is also known as acetate group (e.g. in uvaricin, 92) and R2 could be also a hydrogen instead of the hydroxyl group. Furthermore, the general structure contains two fatty acid chains (depicted in green). The THF moieties were always substituted on positions 2 and 5. The substituent either consists of another THF moiety or a fatty acid chain. One of the fatty acid chains contains an unsaturated γ-lactone (depicted in pink). Moreover, the illustrated hydrogen in the structure of Figure 1, could also be replaced by a hydroxyl group.
Variations are known in the amount of THF units in the THF core, the size of the fatty acid chains, and the hydroxyl groups that could be replaced by hydrogens or acetate groups. Although, a more interesting variation is the changed stereochemistry of the THF core in other Annonaceous Acetogenins.5
Figure 1: The molecular fingerprint of adjacent bis-THF cored Annonaceous Acetogenins
Nomenclature of the bis-THF core
The nomenclature of the core of acetogenins is divided into two groups. First, the number of THF units and whether they are adjacent or non-adjacent. Secondly, the stereochemistry of the THF core is different in several acetogenins. In figure 2, the different acetogenins are depicted having varying stereochemistry’s and/or functional groups. Both (+)-bullatacin and (+)-uvaricin do have an adjacent threo-trans-threo-trans-erythro bis-THF core. (+)-Bullatacin has hydroxyl groups as R1 and R2. However, for (+)-uvaricin R1 is an acetate and R2 is a hydrogen.
6
An important descriptor of the configuration of the bis-THF moieties in acetogenins, is whether they are classified threo or erythro (depicted in red). Since each THF unit has two stereocenters, the relative stereochemistry of the chiral centers describes the stereochemical conformation. Determining the stereochemistry starts from the lowest numbered carbon (so carbon-15 in the top left structure of Figure 2). According to the Fischer projection, the adjacent chiral center could have the oxygen on the same side or in the opposite side. In addition, the stereochemistry in the THF moiety is described as cis when the stereocenters are syn with each other or trans when the stereocenters are trans with each other. The cis-/trans- nomenclature is depicted in blue in Figure 2.
Figure 2: Examples of acetogenins including their stereochemical information of the bis-THF core.
Desymmetrization reactions on C
2-symmetrical molecules
C2 symmetrical objects are able to rotate 180° around its axis, resulting in the same illustration (Figure 3). Many approaches that are discussed in the syntheses, uses convergent synthesis of three building blocks. The preparation of the bis-THF core moiety by total synthesis in the lab could yield in symmetrical building blocks. The approaches require a C2-symemtrical building block since those methodologies gave efficient pathways to the desired building blocks for the total synthesis of the natural products (Figure 3). Since the molecular fingerprint of acetogenins are not symmetrical (as depicted in Figure 1), the symmetry has to be diminished via a desymmetrization step. Here, one side of the core must become attached to a
7
saturated fatty acid chain, but the other end of the building block requires the fatty acid chain containing a γ-lactone. The step where the symmetry is lost is known as desymmetrization (Figure 3).
Figure 3: Explanation of total syntheses using C2 symmetrical building blocks
Biological activity
The interest in Annonaceous Acetogenins is derived from the potent cytotoxic effects against tumor cells.1 The current hypothesis of the cytotoxicity is that the mechanism of action takes place on the mitochondria, which are responsible for the energy production in the cells. Acetogenins promote cellular apoptosis via the inhibition of acetogenins on the mitochondrial and bacterial NADH:ubiquinone oxireductase (complex I). Moreover, the acetogenins affect the electron-transfer step from the high potential iron-sulphur cluster to ubiquinone via acting on the ubiquinone-catalytic site of complex I. Since the electrons cannot enter complex I, the electrons caused a negative accumulation. As result, the inhibition causes blockage of the cells energy production.
According to McLaughlin et al., the γ-lactone (pink in Figure 1) interacts with the mitochondrial complex I. Subsequently, the THF rings flanked by hydroxyl groups functioned as hydrophilic anchors at the membrane surface.5,6
8
Biosynthesis
Polyketide pathway in natural products
The biosynthesis of Annonaceous Acetogenins proceed via the polyketide pathway, which plays an important role in the production of secondary metabolites within bacteria, fungi, plants, and certain marine animals7,8. The polyketide pathway uses acetyl coenzyme A (acetyl-CoA, 1) as a C-2 building block, for the biosynthesis of an array of natural products.9 In the Claisen condensation of acetyl-CoA , a linear carbon chain is formed with one ketone per two carbons (ketone 2). These chains could be reduced by the ketoreductase (KR) enzyme to alcohol 3. In the next step, the dehydratase (DH) enzyme eliminates water to form α,β-unsaturated thioester 4. Subsequently, the enoyl reductase (ER) enzyme is able to reduce the double bond toward thioester 5. This mechanism is illustrated in Scheme 1.9,10
Scheme 1: Fundamental steps in polyketide synthesis: reactions of the biosynthesis of acetogenins
Post-modification(s) in the biosynthesis toward the THF core
The fatty acid chains in acetogenins are fully exposed to three enzymes: KR, DH, and ER. The biosynthesis in the THF core could lack exposures of DH and/or ER for some C2-building block moieties. This results in remaining hydroxyl groups and double bonds at the carbon chain after the polyketide synthesis. The double bonds are epoxidized in a post-modification. There are two possible pathways for epoxides to react as electrophile by the formation of the THF cores:10,11 Both pathways are illustrated in Figure 3.
1. Formation by a hydroxy-group close to the epoxide (-CH2CH2- in between). The aspartic acid amino acid in the enzyme deprotonates the alcohol and tyrosine can stabilize the deprotonated alcohol. The formed alkoxy-group can receive a proton by protonated aspartic acid.
2. Since a lot of water is present, deprotonated aspartic acid in the enzyme can deprotonate water on a similar fashion as the alcohol in the first pathway. The formed hydroxide ion is stabilized by hydrogen bonds via tyrosine and asparagine and can react as nucleophile on epoxides. The formed alkoxy-group can receive a proton by protonated aspartic acid.
9
Figure 3: Illustration of nucleophiles in biosynthesis of THF-fragments
The biosynthesis toward mono-THF cores
The proposed mechanism of the biosynthesis of mono-THF core acetogenins is depicted in Scheme 2. ER exposure would not occur, so the double bonds are retained as illustrated in diene 6. The diene is oxidized to epoxides in a post-modification to obtain bis-epoxide 7. Due to the highly chiral environment of the active site, enzymes are able to perform these reactions in an enantioselective manner, which control the cis- or trans-configuration in the THF-cores.10,12 The hydroxy ion (Figure 3, pathway 2) reacts on the carbon next to R1 or R2 to obtain the desired THF-core 8.
Scheme 2: Mechanism of the biosynthesis of the mono-THF core of acetogenins
The biosynthesis toward adjacent bis-THF cores
Cole and co-workers published the first isolation of an Annonaceous Acetogenins. In addition to that, they proposed a biosynthetic route in the synthesis of uvaricin (Scheme 4).2 The precursor produced from the polyketide synthase is triene 13. Afterwards, triene 13 is enzymatically epoxidized toward tris-epoxide 14. Acetic acid initiates the cascade cyclization to the adjacent bis-THF core 15, as depicted in Scheme 4.
10
None of the proposed mechanisms have been proved, however to our current knowledge on the synthesis and intermediates of polycyclic ethers they provide a reasonable proposition.10,13,14
The biosynthesis toward tris-THF cores and non-adjacent bis-THF cores
Besides the synthesis of the acetogenins containing a mono-THF core, acetyl Co-A is also used in the biosynthesis toward tris-epoxide 10, which serves as a precursor for the tris-THF cores (11) and the non-adjacent bis-THF cores (12), as depicted in Scheme 3. In the biosynthesis toward the precursor, one hydroxy group remains, because the DH-enzyme is not available. Besides that, the three double bonds are obtained from this polyketide synthase. A post-modification oxidizes triene 9 toward tris-epoxide 10. In the synthesis toward the tris-THF-core 11 (mechanism in green), all THF cyclization mechanisms happen in an intramolecular cascade. Alternatively, if water acts as nucleophile, the non-adjacent bis-THF core 12 (mechanism in blue) is synthesized.10
11
Adjacent bis-THF core
Several methodologies have been developed toward the bis-THF core of acetogenins.5,15,16 The most popular methods include (cascade) reactions on epoxides, metal catalyzed reaction and desymmetrization steps. Cascade reactions containing epoxides are “bioinspired” as illustrated in the section of Biosynthesis.
Cascade methodologies
Many bioinspired cyclization reactions and metal catalyzed reactions are known. The last interesting strategy is iodoetherification, due to its usage in both adjacent bis-THF cores and non-adjacent bis-THF cores.
Double S
N2 strategies
This strategy is used in many studies and could be divided in two categories: Inside-Out Cascade Cyclization and Outside-In Cascade Cyclization (Figure 4), as explained by van Lint et al (2019).17
Figure 4: inside-out cascade cyclization (A) and outside-in cascade cyclization (B)
Inside-Out Cyclization
The first synthesis of this THF building block was published by Hoye et al. (1985).18 This approach was based on a tris-epoxide precursor 18 (Scheme 5). The precursor is synthesized in 5 steps from 16, after epoxidation of alkene 17. The most likely place where the hydroxide could react is on the cis epoxide. From there on, the nucleophilic oxygens can react on the epoxides on an Inside-out Cyclization as the blue arrow in Figure 4A. Then the formed alkoxy-ion can react in a cascade fashion where the acetate acts as leaving group (LG). The desired adjacent bis-THF core is synthesized as compound 19.
Scheme 5: The synthesis of a bis-THF building block
The drawback of this methodology is that a diol could be formed as well, which undergo a Payne rearrangement as illustrated in Figure 5. This side reaction was favored according to the low enantiomeric access measured by Hoye et al. Besides that, another drawback of this methodology is the C2-symmetry. Two the same hydroxyl groups (red in Figure 5) must be replaced by different building blocks. Besides
12
that, two other hydroxyl groups (pink in Figure 5) are present in this building block. The authors already concluded there is improvement possible toward the synthesis of the adjacent bis-THF building block, due to the racemization in this approach. Racemization happens because of the many electrophilic positions in precursor 18. Possible unwanted products are illustrated in Figure 5. Furthermore, the other hydroxyl group of 18 can undergo Payne rearrangement as well.
Figure 5: Payne rearrangement
The racemization has been solved by a stereocontrolled approach, as published by Hoye et al. one year later.19 In this approach, precursor 21 underwent an inside-out cascade cyclization under acetonide removal in acidic medium. This approach is illustrated in Scheme 6.
Scheme 6: Stereocontrolled Inside-out Cyclization of Hoye et al.
Starting from commercially available 1,4-di-O-tosyl-2,3-O-isopropylidene-L-threitol (20), the desired precursor (21) can be obtained by an yield of 4% to 14%. The authors did not report the yields of the Sharpless epoxidation but referred to previous and similar syntheses. Treatment of precursor 21 with H+, THF, H2O result in the desired bis-THF core 22.
Other synthetic procedures used these inside-out cascade cyclization in the synthesis to a C2-symmetrical building block for total synthesis.20,21 Two total syntheses are described are described in Scheme 7.
13
In the first total synthesis, Scheme 7 illustrates the use of this mechanism in the total synthesis toward (+)-15,16,19,20,23,24-hexepi-Uvaricin20 (Total Synthesis A) and (+)-Parviflorin21 (Total Synthesis B).
Scheme 7: Total syntheses of the Inside-Out Cyclization by Hoye
The synthesis toward (+)-15,16,19,20,23,24-hexepi-uvaricin (26) is performed starting from dioxolane 23. Precursor 24 was synthesized in five steps from bis(iodomethyl) dimethyl-dioxolane 23. Desymmetrization of 24 was achieved by installation of paratoluene sulfonyl (Ts) group and the cyclization toward the bis-THF ring system is catalyzed by Amberlyst-15, providing compound 25. bis-THF core 25 is used as building block in the total synthesis toward a stereoisomer of uvaricin. The desired product, (+)-15,16,20,21,23,24-hexepi-uvaricin (26), was synthesized in 13 steps from building block 25 in less than 13%.20
The second total synthesis that is depicted In Scheme 7 started from triene 27. The triene was converted to 28 in four steps with an overall yield of 44%. Protection of the alcohols by TBDPS is followed by an asymmetric dihydroxylation on the remaining double bond. Under acidic conditions, the desired THF-core 29 has been synthesized with a yield of 85% (both steps from 28 toward 29). From THF core 29, the total synthesis toward (+)-Parviflorin (30) has been completed by 8 more steps with an overall yield (over these last 8 steps) of 22%.21
Application of this strategy, based on the dioxolane core of precursors 21 and 24, can be found in several reports on the synthesis of the bis-THF cores of natural product (+)-Bullatacin and numerous acetogenin mimics.22,23
14
Additionally, van Lint et al. (2019) described the synthesis toward a new C2-symmmetrical building block as key intermediate in the envisioned total synthesis multiple acetogenins, e.g. bullatacin and rolliniastatin (Scheme 8).17
Scheme 8: Building block synthesis by van Lint et al. (2019)
In Scheme 8, commercially available alkyne 31 was converted in 10 steps to precursor 32 in an overall yield of 9%. The cascade cyclization toward bis-THF 33 occurs after deprotection of the benzaldehyde acetal, under hydrogenation conditions. Using basic conditions, C2-symmetric bis-THF-bis-epoxide 34 was obtained.
Besides the approaches illustrated in Scheme 1 to 8, Marshall et al. introduced a new approach to inside-out cascade cyclization reactions. The proposed precursor contains an internal nucleophile: the silyl protected hydroxy groups which act as nucleophile after deprotection (Scheme 9).24
15
In this approach, diester 35 could lead to four different precursors (36, 38, 40, and 42) in four steps. The observed yields vary from 26% to 67%. In all cases, deprotection of the hydroxy groups initiates a cascade cyclization reactions toward the desired adjacent bis-THF core (37, 39, 41, and 43), which could be used as building block in total syntheses.24 Although there are many other possibilities to control the stereoisomeric ratio of products using chiral tartrates, the approach illustrated in Scheme 9 shows an expansion on the synthetic possibilities. The THF core building block from this approach is used for the total synthesis toward (+)-Asimicin25, (30S)-Bullanin26 and Asimin.27
Another Inside-Out Cyclization has been published by Li et al. in 1999. Again, featuring the alkoxy nucleophile, obtained after TBS deprotection. However, in this approach, the sp2-hybridized atom acts as electrophile. In this approach, the generated THF building block has double bonds adjacent to THF-moiety (Scheme 10).28
Scheme 10: Substitution reaction of Li et al. (1999)28
As illustrated In Scheme 10, cyclic diene 44 was first mono-epoxidated using m-CPBA after which hydrolysis with lithium hydroxide afforded the corresponding diol. The hydroxy groups were protected as a silyl enol ether and the remaining alkene was subsequently subjected to ozonolysis, affording bis-aldehyde 45. Route A occurs via a Wittig reaction and is reduced by DIBAL. In the last step, precursor 46 is obtained by an Appel reaction. The cyclization is performed by removal of the TBS protecting group. Under basic conditions, the cascade reaction is promoted toward the desired bis-THF core 47 and 48 in a 2:1 ratio (47:48). Since the authors encountered the low selectivity toward stereoisomer 47 route B was described, to access precursor 46.
A challenge in these reactions listed above is the treatment of these building blocks in the next steps. The building blocks (19, 22, 25, 29, 33, 34, 37, 39, 41, 42, 47, 48) all are C2-symmetrical, so a desymmmetrization step is required to produce acetogenin natural products. The desymmetrization step is challenging because one functional group must react, while two the same of these functional groups are available. From a practical perspective, a double reaction on both functional groups is hard to avoid. A way to avoid this challenge is published by Marshall et al. (2003).29 This approach is illustrated in Scheme 11 where different electrophiles are used.
16
Scheme 11: Total synthesis toward Asiminocin without a desymmetrization step
The total synthesis is based on the synthesis of three building blocks. The first building block is synthesized from alcohol 49 to building block 50. The second building block is synthesized from hemiacetal 51 toward building block 52 and the third building block is synthesized from alkyne 53 toward building block 54. Building blocks 50 and 52 are coupled via nucleophilic substitution to the carbonyl of building block 50. Subsequently, the benzyl protected hydroxyl group is replaced by an aldehyde. This reaction is followed by the coupling of building block 54 via nucleophilic substitution to the carbonyl to obtain precursor 55. Precursor 55 undergoes inside-out cascade cyclization after TBS deprotection toward the desired bis-THF core 56. Bis-THF core 56 is used for the total synthesis of asiminocin.29
Besides the strategy depicted in Scheme 11, desymmetrization steps are also avoided when performing the cyclization reaction in two separated steps.30 This approach is illustrated in Scheme 12. Bromide 57 is converted to diol 58 in 10 steps with an overall yield of 35%.31 From diol 58, the precursor 59 could be obtained in 9 steps (37% overall yield). Deprotection of the TBS group lead to the formation of the diol instead of the desired bis-THF moiety (96% yield). To obtain the desired building block 60, a introduction of an acetonide protecting group was needed (93% yield) and the epoxide was synthesized after K2CO3 treatment (99% yield).30 This building block is used in total synthesis of 15-epi-Annonin I (61). From 60 the desired natural product of 61 is obtained in 9 steps with an overall yield of 19%.
17
Outside-In Cyclization
Besides the “inside-out cascade cyclization” mechanism, the outside-in approaches are also known (Figure 4B). Van Lint et al. (2019) published that this approach would lead to another stereoisomer than illustrated in Scheme 8. Scheme 13 illustrates the approach toward erythro-cis-threo-cis-erythro bis-THF rings.17
Scheme 13: Outside-In Cascade Cyclization of van Lint et al. (2019)
Alkyne 31 was converted to diol 62 in 9 steps with an overall yield of 15%. Diol 62 undergoes cascade cyclization under basic conditions, yielding bis-THF moiety 63 in a yield of 66%. After Payne rearrangement the alkoxy ion displaces (SN2) the tosylate groups. The desired bis-THF core 63 is obtained with a yield of 66% from precursor 62.
The outside-in cyclization reaction occurs separately for avoiding desymmetrization steps. An example of this methodology is illustrated in Scheme 15. Diol 64 is obtained from kinetic resolution of racemic 1-tetradecene oxide with Jacobsen’s catalyst32,33. The synthesis toward precursor 65 continued that acts as precursor for the first ring cyclization.34 The next THF moiety is formed from 66, which is obtained in 4 steps with 69% overall yield from 65. Cyclization to the desired THF-cored building block 68 is obtained in 77% yield. In this synthesis toward the THF building block, also the fatty acid chain R2 (Scheme 2) is already introduced.35
Scheme 14: Separated outside-in cascade cyclization (both steps), including saturated fatty acid chain (R2)
Metal catalyzed strategies
Metal catalyzed reaction strategies provide efficient possibilities for the synthesis toward several bis-THF moieties. In chapter 2.1.1 (Double substitution strategies), numerous reports were discussed. Most of these articles only provide limited possibilities. For example, the high amount of papers about trans configured THF rings, while just a few examples are known about the synthesis toward the cis-THF ring. The reactions in this chapter provide an expansion on the possibilities discussed in the double substitution strategies.
Rhenium oxide catalysis
Sinha et al. (1995) published a new strategy to synthesize (intramolecular) THF units. The discovery was based on the tandem oxidative cyclization reaction using Re(VII). Both Re2O7/lutidine and Re2O7/H5IO6 worked well in the mono and double cyclization illustrated in Figure 6.36 Besides that, CF3CO2ReO3 was used for the cyclization of tris-THF units.4
18
Figure 6: Discovered routes by tandem oxidative cyclization using rhenium oxide of Sinha et al.4,36
Cis-THF ring are formed in the reactions illustrated in Figure 6. However, when the precursor of these reactions are trans alkenes, the opposite stereochemistry is obtained.37
Total syntheses toward asimicin and bullatacin are published by Avedissian et al. (2000). The total synthesis requires a THF building block (X, illustrated in Scheme 15). From building block 73, asimicin is synthesized in 3 steps with a yield of 41%.38
In the approach A (Scheme 15), aldehyde 69 is converted to lactone 70 by 9 steps in 20% overall yield. Afterwards, cyclization by rhenium oxide occurs, yielding alcohol 71 in 64% yield. Subsequently, diol 72 is obtained in two steps from alcohol 71 in 85% overall yield. A SN2 substitution at diol 72 leads to the desired building block 73 in 81% yield. Approach B starts from alkyne 74, which is converted to ester 75 by 12 steps in an overall yield of 24%. Ring formation occurred via cascade SN2 substitution toward alcohol 76 in 43% yield. Afterwards, the next cyclization occurred by a rhenium oxide catalyzed reaction, obtaining building block 73 in 69% yield. Approach C was performed starting from diene 77, which was converted to alkyne 78 by 6 steps in 21% overall yield. Cyclization occurred by a rhenium oxide catalyzed reaction toward alcohol 79 in 68% yield. Alcohol 79 is used to synthesize ester 80 in 4 steps with a yield of 38%. In addition to the first cyclization, the second cyclisation was performed by rhenium oxide to obtain building block 73 in 69% yield.
19
Scheme 15: Controlled building block synthesis for the total synthesis toward asimicin.
Furthermore, the synthetic strategy has been used to the syntheses of more types adjacent bis-THF cores.39
Cobalt, bis(1-morpholinocarbamoyl-4,4-dimethyl-1,3-pentanedionato)cobalt(II)
Cobalt(II) complexes are able to convert hydroxy-alkenes into THF-units.40 The approach is used in the syntheses toward (-)-Asimilobin and (+)-Asimilobin (Scheme 16).41,42 In the Scheme, only the syntheses toward (-)-Asimilobin is illustrated, where (+)-Asimilobin is obtained from the opposite dihydroxylation in the first step from 81.
20 Scheme 16: Total syntheses of (-)-Asimilobin.
The reaction started with a regioselective asymmetric dihydroxylation of triene 81 toward diol 82. Diol 82 can cyclize in addition of a Cobalt(II)-complex, to obtain THF Core 83. The desired THF Core 83 can be converted to (-)-Asimilobin (84) in 10 steps with an overall yield of 4% (from 83).
The step from triene 81 toward 82 was a changed approach compared to an approach published by Wang et al. in 1999.43 In this approach, alkene 85 is used as starting material and is converted to diol 82 in 3 steps (66%).
Palladium(II) catalysis
Palladium complexes act as catalyst for the formation of tetrahydrofuran from vicinal diol allyl acetates.44,45 Burke et al. (2001) used the knowledge for the formal synthesis of uvaricin, which is depicted in Scheme 17.46 Diethyl D-tartrate (86) was used as starting material for the synthesis toward diol 87, which was performed via 7 steps in an overall yield of 44%. Diol 87 is used as precursor for the cyclization reactions. The use of nonchiral ligands gives a mixture of 88, 89, and 90 (1:2:1). However, Trost et al. developed chiral ligands as N-[2-(2′-diphenylphosphino)benzamino-cyclohexyl](2′-diphenylphosphino)benzamide (DPPBA). The use of these chiral ligands gave only one type of building block. (R,R)-DPPBA resulted in building block 88, which was slightly modified in 3 steps to epoxide 89 (56% yield). Epoxide 89 is used as THF core for the total synthesis toward uvaricin (92). Building block 90 is obtained from the cyclization reaction with (S,S)-DPPBA as ligand, which could be used In total syntheses toward rollimembrin and membrarollin.
21
Scheme 17: Palladium catalyzed reactions to adjacent bis-THF cores
In addition to the total synthesis of uvaricin, the palladium catalyzed cyclization is used in the total synthesis of (+)-Goniocin. However, in this total synthesis no desymmetrization step was required due to step by step installation of the bis-THF fragment. The first part of the total synthesis is depicted in Scheme 18. The entire total synthesis is omitted from scope, since the natural product contains a tris-THF moiety instead of a bis-THF moiety. The total synthesis started from epoxide 93. The desired precursor 94 was obtained in 2 steps (68% overall yield). The Pd(II)-catalyzed cyclization was performed to obtain mono-THF core 95. In 6 steps, the precursor was synthesized for the second cyclization. The precursor was obtained via alcohol 96 (in 5 steps from mono-THF core 95, 62% yield). The Grubbs metathesis was performed and followed by the second ring cyclization. Bis-THF moiety 97 was obtained from alcohol 96 with an overall yield of 72%.47
Scheme 18: First part of the total synthesis of (+)-Goniocin (untill the bis-THF fragment)
22
Uenishi et al. (2011) published a Pd(II)-Catalyzed triggered cyclization by epoxides to obtain non-symmetrical building blocks in one cascade reaction. As illustrated in Scheme 17 from the previous example, THF-ring formation is favored rather than THP-ring formation. The same rule is applied for the mechanism depicted in Scheme 19.48 In this example, the epoxide 98 is converted to bis-THF fragment 100 via intermediate 99.
Scheme 19: Mechanism of Palladium catalyzed triggered cyclization of epoxy, hydroxy allylic alcohols The mechanism is used to synthesize bis-THF moieties on a stereospecific manner.
The final palladium catalyzed approach has been published by Daniels et al. in 2011. In this approach, 7-membered cyclic carbonates were treated with Suzuki Cross Coupling conditions. In the proposed mechanism, palladium can coordinate to the alkyne moiety, so CO2 of the carbonate left the complex. Then, the formed sp2-hydribised carbon allows electrophilic attack of alcohol to obtain a THF moiety, as illustrated in Figure 7.49
Figure 7: Proposed mechanism of THF ring formation
As result of this observation, the authors realized that a carbonate moiety and a hydroxyl group must cyclize to obtain cyclic ethers. The authors started with a general methodology to illustrate the approach of this reaction. The general methodology is known as the “state-of-the-art” depicted in Scheme 20. Both cyclic carbonates (using 101) as acyclic carbonates (using 102) yields in THF moiety 103. In the synthesis toward bis-THF cores, two of these fragments (e.g. 102) must be attached to each other. In that approach, diene 104 was selectively converted to precursor 105 in 7 steps with an overall yield of 15%. Then the cyclization toward bis-THF moiety 106 was performed. Various stereoisomers of precursor 105 gave different yields. The desired bis-THF core was obtained in yields of 71% to 82% from 105.
23
Scheme 20: Daniels et al. (2011) approach using Pd(II) for the syntheses of bis-THF moieties.
Ruthenium oxide catalysis
Ruthenium oxide (RuO4) can perform cyclization reactions on farnesyl acetate (107) or geranylgeranyl acetate. RuO4 can coordinate to 5 membered metallocyclic species (108) as depicted in Scheme 21. Subsequently, the other double bond coordinates to RuO4 to obtain acetate 109. A second [3 + 2] cycloaddition leads to acetate 110. Cleavage of the ruthenium oxide happens via two possible pathways. The first pathway is illustrated in green in Scheme 21. Oxidative cleavage could happen to obtain bis-THF core 111. Furthermore, in the second pathway hydrolysis must occur to obtain bis-THF core 112. Under ruthenium oxide conditions, oxidation of bis-THF core 112 leads to bis-THF core 111.50
Scheme 21: Mechanism of THF formation by ruthenium oxide catalysis
This approach is used in the synthesis of new bis-THF cored building blocks. Important to note in this approach is that only cis-THF rings are synthesized due to the chelate effect of the ruthenium oxide.51 Due to the E-geometry of the double bonds, threo stereochemistry is obtained in between the THF rings (see Scheme 22). Starting from undecanal (113) triene 114 is obtained in 11 steps with an overall yield of 9%. NaIO4 is added to oxidize ruthenium oxide. Less NaIO4 would lead to more hydrolysis, obtaining a larger fraction of 116 (green box), while more equivalents of NaIO4 gives more oxidized bis-THF ring moieties (115 and 117, brown boxes). NaIO4 promotes rather oxidation than hydrolysis due to capacity of NaIO4 to generate oxidations. When too small amounts of oxygen are present, the hydrolysis side reaction takes is favored over the oxidation mechanism.
24
Scheme 22: Ruthenium oxide used in the synthesis of adjacent bis-THF cored building blocks
A new methodology to synthesize precursors for ruthenium oxide catalysis is published by Grubbs and co -workers (2016). However the authors only synthesized precursors for mono-THF cores, which is out of the scope of this literature thesis.52
Iodoetherification
Iodoetherification is a strategy where THF units can be synthesized by using several iodide donors. In this strategy, a nucleophilic oxygen is needed that is able to attack a double bond with coordinated iodide. An example is illustrated in the figure 8, where the THP unit is converted to a THF unit.53 The E/Z geometry of the double bond determines the stereochemical outcome of the iodide.
Figure 8: Mechanism of iodoetherification according to Zhang et al. (2002)
A more general approach in iodoetherifications toward acetogenin THF cores is depicted in Figure 9. The hydroxy can act as nucleophile toward the activated double bond. The obtained product is a trans-THF ring. However, a cis-THF ring could be obtained when a silyl protected alcohol or ether is used instead.54
Figure 9: General mechanism of iodoetherification
Starting rom aldehyde 118 (Scheme 23), the synthesis toward a bis-THF cored building block was prepared. The precursor (119) was synthesized in one step with 65% yield after flash column separation of both enantiomers. Subsequently, the iodoetherification was performed in order to perform a ring cyclization toward bis-THF moiety 200 (72% yield). The cyclization results in both enantiomers of the cis-trans bis-THF ring and the cis-trans-cis-trans bis-THF ring.55
25 Scheme 23: Racemic synthesis of bis-THF core
Although iodoetherification strategies do not seems to be diastereoselective, Ruan and Mootoo were able to publish a formal syntheses to trilobacin and asimicin in 1999.56 The iodoetherification was based on a different reagent: iodonium dicollidine perchlorate (IDCP), which acts as strong I+-donor.57 as illustrated in Scheme 24. Starting from alkene 121, the desired precursor 122 was obtained in 5 steps with an overall yield of 33%. The first cyclization was performed to obtain bis-acetal 123 in 81% yield. Subsequently, oxidation and Wittig resulted in alcohol 124. The second ring cyclization was performed via a SN2-substitution reaction. In the blue box toward subunit 125, the iodide was replaced by another leaving group with the opposite stereochemistry. In this case an asimicin subunit was prepared. By using the iodide as leaving group (green box toward subunit 126), the trilobacin subunit was synthesized.
Scheme 24: Formal synthesis toward asimicin subunit 125 and trilobacin subunit 126
Epoxide as nucleophile
This section shows the last type of cascade reactions, although sometimes the strategy cannot happen in a cascade manner. The reason behind this is that both articles discussed in this session perform their cyclization reactions separately. One of the cyclization reactions is the creation of an epoxide that react directly further to a THF unit, while asymmetric dihydroxylation after the installation of the epoxide require an additional step (Scheme 26).58,59
The total synthesis toward trilobin (132) is depicted in Scheme 25. The total synthesis started from alcohol 127. Acetonide 128 was obtained from alcohol 127 in 8 steps with an overall yield of 39%. In the
26
installation of the first THF-moiety from acetonide 128 to lactone 129, acetonide hydrolysis was performed. Afterwards, basic conditions were used for the cyclization toward lactone 129. Then, precursor 130 was prepared for the installation of the second THF-moiety. The preparation toward precursor 130 was performed in 4 steps with a yield of 76%. Epoxidation on the double bond resulted in a cascade reaction toward bis-THF core 131. Trilobin (132) was prepared in 12 steps from building block bis-THF core 131, with a yield of 11%.58
Scheme 25: Total Synthesis of Trilobin
The other total synthesis is illustrated in Scheme 26. In the total synthesis, Squamotacin (138) is obtained. Starting from ester 133, precursor 134 is obtained in 11 steps with an overall yield of 28%. The first ring cyclization was performed in a non-cascade manner toward lactone 135. Asymmetric dihydroxylation and catalysis by Acidic Amberlyst gave lactone 135 with 64% yield. Four steps are needed to convert lactone 135 into diol 136. Diol 136 undergoes ring cyclization after epoxidation on the double bond (on the same manner as in Scheme 25 from 130 to 131). The desired bis-THF core is obtained and used as building block for the further total synthesis of squamolacin. Seven steps with a yield of 15% are needed to convert bis-THF building block 137 into squamolacin (138).59
27
Non-Cascade methodologies
Moreover, non-cascade methodologies have been published. Most of these articles were published close or shortly after the millennium.
Double S
N2 strategies
These methodologies are similar as the “Cascade Double SN2 strategies” (Chapter 2.1.1). For example, in Scheme 9, 11, or 14 TBS deprotection led to the electrophilic alkoxy group. Those species will undergo cyclization in a cascade manner. However, in the non-cascade double SN2 strategies, another step is required for the cyclization.
Inside-Out Cyclization
For the inside-out approach, a total synthesis toward 14,21-diepi-squamocin-K (142) was published.60 The total synthesis is depicted in Scheme 27. The synthesis starts from diol 139. Three equivalent diol 139 moieties are required for the preparation of precursor 140 over 8 steps in an overall yield of 10%. Subsequently, the hydroxyl groups must be activated as leaving groups and the acetonide must be hydrolyzed. Then, the cyclization toward bis-THF building block 141 takes place under basic conditions (59% yield over 3 steps). Building block 141 is used for the total synthesis toward 14,21-diepi-squamocin K (142) by 6 steps with a yield of 23%.60
Scheme 27: Total synthesis of 14,21-diepi-squamocin-K
Outside-In Cyclization
In addition to the inside-out cyclization, approaches with the outside-in cyclization are known. For example, the synthesis toward building block 145 (Scheme 28 that could be used for total syntheses toward (+)-Parviflorin, (+)-Squamocin K, and (+)-5S-Hydroxyparviflorin. The synthesis started from alcohol 143. Alcohol 143 has been converted toward precursor 144 in 8 steps with an overall yield of 23%. The acetonide groups are hydrolyzed using Amberlyst and fully cyclized under basic conditions toward the desired THF cored 145 (61% yield).61 Afterwards, five steps are required to synthesize (+)-parviflorin (146) with an overall yield from 145 of 35%.
28 Scheme 28: Approach of Trost et al. (1997)61
In the total synthesis to uvaricin (92) as depicted in Scheme 29, undecanal (113) was converted to precursor 143 in 16 steps with an overall yield of 12%. Acetonide hydrolysis of precursor 143 is followed by cyclization by basic conditions. Although deprotected bis-THF moiety 144 was synthesized (13%), also the MOM-protected moiety was obtained with a yield of 48%. The protected bis-THF moiety was deprotected by an additional step with 79% yield. Five more steps were required to obtain the desired bis-THF building block 145 which was required for the total synthesis. Uvaricin (92) was synthesized in 4 steps from building block 145 with 9% overall yield.62
Scheme 29: Total synthesis toward uvaricin
Emde and Koert published syntheses toward multiple Annonaceous acetogenins via building block 149, depicted in Scheme 30. The synthesis starts from 146 that is converted to precursor 147 in 5 steps with 50% overall yield. Subsequently, bis-THF moiety 148 is synthesized under basic conditions and obtained with 92% yield. The desired building block 149 is obtained in 2 steps from 148 with a yield of 39%. From building block 149, both squamocin A (150) and squamocin D (151) were obtained in 16 steps with yields of 1% and 3% respectively.63,64
29
Scheme 30: Synthesis toward building block 149 in the approach of Emde and Koert
Another approach using double SN2 outside-in cyclization is published by Nattrass et al. in 2005. The total synthesis is based on acetal hydrolysis, however differs from the previously described method by Trost et al.61 and Emde et al.63,64, as depicted in Scheme 28, 29 and 30. The total synthesis is depicted in Scheme 31 and starts from (S,S)-dimethyl-L-tartrate 152. 13 steps are required to synthesize precursor 153 in 19% overall yield. Subsequently, acetal hydrolysis was followed by a basic treatment, so that the precursor underwent the desired cyclization toward bis-THF fragment 154. Finally, 12 steps are needed to synthesize 10-Hydroxyasimicin (155) from bis-THF building block 154 (7% yield).65
Scheme 31: Outside-in cyclization from the approach of Nattrass et al. (2005) toward 10-Hydroxyasimicin The last synthesis is based on the same an α,β-unsaturated ester which acts as precursor. The synthesis toward building block 158 is depicted in Scheme 32 and starts from alcohol 156. Alcohol 156 is converted to precursor 157 that underwent cyclization after asymmetric dihydroxylation and treatment under basic conditions. The building block Is not used in total syntheses, but in the synthesis of Annonaceous acetogenin analogs.66,67
Scheme 32: Synthesis of building block 158
30
Single Substitution strategies
In addition to double SN2 strategies, single substitution reactions are performed as well. Many of these reactions were used in combination with other strategies. However, the most important will be treated in this chapter.
Epoxides as electrophile
According to Scheme 15, only Approach C has been fully catalyzed by rhenium oxide for both ring cyclization reactions. In these type of syntheses toward building block 73, epoxides are used as well as electrophiles for THF-ring cyclization.68 Besides that, as expansion to the cascade strategies discussed in Chapter 2.1.3, also non-cascade strategies are known where epoxides acts nucleophile. For example in the synthesis toward squamotacin, from precursor 134 to lactone 135.59
A strategy for the synthesis of bis-THF moieties via non-cascade single substitution is published by Bertrand and Gesson in 1992.69 This approach uses epoxides as electrophile and is depicted in Scheme 33.
Scheme 33: Racemic single substitution strategy by using an epoxide as electrophile by Bertrand et al. From aldehyde 159, alkene 160 is synthesized in two steps. Afterwards, cyclization toward both bis-THF core 161 and bis-THF core 162 was performed by treating the epoxide with acid. The reaction gives both diastereomers, due to the epoxidation occurs on both sides of the molecule.
In addition, epoxide 164 in Scheme 34 perform a cyclization reaction via single substitution on epoxides. Epoxide 164 is made from glutamic acid (163). Subsequently, cyclization occurred to afforded bis-THF building block 165. From building block 165, 16 steps were required to synthesize (+)-Rolliniastatin 1 (166) in 10% overall yield.70
Scheme 34: Stereospecific single substitution strategy using an epoxide as electrophile
Leaving groups as electrophile
Single substitution strategy by using an leaving group as electrophile is used in the total synthesis toward rolliniastatin 1, rollimembrin, and membranacin.71 The total syntheses are based on building block 172 (Scheme 35) that is synthesized from diol 167. Diol 167 was used to synthesize precursor 168 by 10 steps
31
in 21% overall yield. The ring formation toward acetal 169 is synthesized in 93% yield and followed by 6 steps toward precursor 170 in an overall yield of 24%. The second cyclization was performed toward bis-THF moiety 171 in 88% yield. Afterwards, two additional steps are required to obtain the desired bis-bis-THF building block 172 in 76% yield. Although building block 172 was synthesized stereoselective, low yields were observed for two steps in the synthesis from 169 to 170. Synthesizing building block 172 was also accessible from alcohol 173. The approach starting from alcohol 173 needed 7 steps to obtain precursor 174 in an overall yield of 42%. Subsequently, the first ring formation was performed, to obtain benzyl ether 175 in 95% yield. Benzyl ether 175 was converted into precursor 176 by 6 steps in an overall yield of 48%. The second ring cyclization was performed to obtain bis-THF moiety 177 in 95% yield. The reaction was followed by 8 steps to obtain the desired bis-THF building block 172. From building block 172, four more steps were required for the syntheses toward rolliniastatin 1 (166), 58% overall yield), rollimembrin (49% overall yield) and membranacin (55% overall yield).
32
Non-adjacent bis-THF Core
Besides the natural occurrence of acetogenins with adjacent bis-THF cores, a variety of acetogenins exist having two non-adjacent THF cores. Most syntheses toward the bis-THF core are based on a convergent synthesis using two building blocks. However, strategies without the use of these building blocks are also known, where a linear route toward the bis-THF core is used in the synthetic approach. This chapter is divided in these methodologies, where in the first part linear synthetic approaches are discussed. Subsequently, in the second part the convergent synthetic approaches using two building blocks toward the bis-THF cores are discussed.
Without the use of building blocks
The first total synthesis was published by Makabe et al. in 1997 (Scheme 37).72,73 The synthesis started with bromide 181. However the authors were able to obtain precursor 182, that was obtained in a diminished overall yield of less than 1.74,75 Cyclization was performed via three steps: mesylation of the single hydroxyl group, asymmetric dihydroxylation of the alkene followed by Triton B facilitated cyclization (48% over 3 steps). From there, three steps were required to synthesize (+)-4-Deoygigantecin (184).
Scheme 37: Total synthesis of the non-adjacent bis-THF Annonaceous acetogenin (+)-4-Deoxygigantecin In addition to the approach of Makabe et al., a procedure toward squamostatin-D was published by Marshall et al. (190, Scheme 38).76 The THF ring formation is based on a similar strategy as discussed in the cascade strategy (chapter 2.1.1, Scheme 9) toward adjacent bis-THF cores using silyl enol ethers as TBS. Precursor 186 is prepared in 6 steps from ester 185 in 84% overall yield. The precursor underwent cyclization toward mono-THF moiety 187 directly after TBS deprotection. The second THF ring was installed from precursor 188, which was prepared in 10 steps with an overall yield of 28%. The installation was also performed by TBS deprotection toward 189. In the last four steps, squamostatin-D was obtained in a yield of 71%.76
33
Scheme 38: Total synthesis toward squamostatin-D in the approach of Marshall et al.76
The third approach is based on [(Trimethylsilyl)oxy]furan (191), which is a useful molecule for installing THF moieties in molecules.77 As depicted in Scheme 39, a formal synthesis was proposed toward the non-adjacent bis-THF cored (-)-4-deoxygigantecin precursors (197 and 198).78
Scheme 39: Formal synthesis toward (-)-4-deoxygigantecin precursors
Lactone 193 is prepared from arene 191 and aldehyde 192 by three steps in 78% yield. The lactone was reduced using DIBAL-H and transformed into THF moiety 194 . The acetate acts as good leaving group, that is replaced by another [(Trimethylsilyl)oxy]furan molecule, obtaining lactone 195. Ring opening must take place to create the desired linker in between the THF moieties of non-adjacent bis-THF cores. The ring opening was performed enantioselective with 8 steps in 28% overall yield, obtaining aldehyde 196. The next [(Trimethylsilyl)oxy]furan molecule (191) was installed on the same manner as from arene 191 and aldehyde 192 toward lactone 193. After reducing, both lactone 197 and lactone 198 are obtained as 1:1 mixture in an 40% yield from aldehyde 196.78
The last approaches toward non-adjacent bis-THF cored molecules require a C2-symmetrical building block. Both approaches are illustrated in Scheme 40. Both approaches started from tetradecatetraene 200, which is obtained from cyclic triene 199 with three steps in 39% overall yield. In Approach A, the double bonds are transformed into acetonide moieties and further treatment on the external double bonds led to diene 201 with four steps in 18% overall yield.79 Since the authors realized that another intermediate might be useful, another route has been developed via alkene 202 (Approach B).80 Diene
34
201 was used for cyclization reaction using osmium metal catalysis36 toward bis-THF core 203 in 77% yield. The synthesis toward the desired precursor was prepared in four steps from 203, obtaining precursor 204 in 47% overall yield. The building block is used for the total synthesis toward (+)-cis-sylvaticin that is synthesized with three steps from 204 in 54%. Besides that, alkene 202 is used for the synthesis toward diene 205 with five steps in 51% overall yield, which is converted to bis-THF core 206 after cyclization in 76% yield. Subsequently, seven more steps were required to synthesize the desired building block (207) in 25% overall yield. Three more steps were required to complete the total synthesis toward (+)-sylvaticin in 56% yield.79,80
Scheme 40: Total syntheses toward (+)-cis-sylvaticin and (+)-sylvaticin
The use of building blocks
A common methodology in the synthesis of non-adjacent bis-THF acetogenins is the use of building blocks, where two THF units are attached to each other. The state of the art is illustrated in Figure 10, which is used in the first strategy.
Bimolecular diene cross metathesis
The first strategy of synthesis toward non-adjacent bis-THF cores using building blocks are two THF moieties that undergo diene cross metathesis. The specific strategy is depicted in Figure 10.
35
Figure 10: State of the art bimolecular diene cross metathesis obtaining non-adjacent bis-THF cores Figure 10 illustrated the building blocks which were used by Zhu and Mootoo in 2003 for the formal synthesis toward bullatanocin, which is also known as squamostatin C (Scheme 41).81 Both building blocks were synthesized from diene 208, yielding acetonide 209 in 57% overall yield and acetonide 211 in 45% overall yield. Both 209 as 211 underwent cyclization via iodoetherification, which is described as useful technique for cyclization reactions toward THF moieties in this literature thesis (Chapter 2.1.3).56,57 The products that were obtained from acetonide 209 and acetonide 211 after ring formation using iodoetherification were obtained mono-THF 210 (in 89% yield) and mono-THF 212 (in 79% yield) respectively. The mono-THF units 210 and 212 were converted to the building blocks (213 and 214) for Grubbs alkene metathesis. The Grubbs metathesis was performed82 and followed by four more synthetic steps to the desired building block 215, that can be used in the total synthesis toward bullatanocin.
Scheme 41: Formal synthesis toward bullatanocin.
Other synthesis from bimolecular building blocks
Another way to attach building blocks together is reported by Crimmins and She in 2004.83 The approach is based on asymmetric acetylide addition from alkyne 218 and aldehyde 221 (Scheme 42). The synthesis of the first building block (218) started from epoxide 216, which is converted to diene 217 with 8 steps in 54% overall yield. Subsequently, Grubbs metathesis and deprotection of TIPS resulted in building block 218 in 97% yield. The second building block (221) was prepared from epoxide 219 (ent-216), which was transformed into diene 220 with 8 steps in 44% overall yield. Afterwards, Grubbs metathesis, double bond removal and hydroxyl epoxidation to aldehyde was performed to obtain building block 221 in 82% yield. The building blocks were attached by asymmetric acetylide addition (70% yield) toward bis-THF 222 and the total synthesis was completed with 7 steps from bis-THF moiety 222, obtaining the desired (+)-gigantecin (223) in 35% overall yield.
36 Scheme 42: Total synthesis toward (+)-gigantecin
Unimolecular silicon tethered ring closing metathesis
A common methodology applied in the synthesis of these non-adjacent bis-THF core building blocks is the Grubbs metathesis. Grubbs metathesis is used as handle for the attachment of building blocks. However, when multiple double bonds are present in the building blocks (e.g. 226, Scheme 43), the metathesis could happen at both double bonds. Hoye et al. proposed a selective Grubbs metathesis. The vinyl hydroxyl double bonds are available for selective Grubbs metathesis using a silicon tether. As result, the double bonds that are illustrated in red in Scheme 43 underwent a selective silicon tethered ring closing metathesis, while the other double bond (in pink) is available for Grubbs metathesis of another building block.
The first approach using a silicon tether is depicted in Scheme 43. The first building block was synthesized from aldehyde 224, that underwent allylation which was treated with DIBAL-H, followed by iodoetherification toward mono-THF moiety 225. Subsequently, building block 226 has been produced via an Appel reaction. The other building block (228) prepared on a similar manner starting from lactone 227. Both building blocks (226 and 228) were treated with Ph2SiCl2 to obtain unimolecular molecule (229 in 52% yield), that is able to perform the Grubbs metathesis on a selective manner. Besides that, the lactone fragment is installed in a one-pot fashion to obtain bis-THF moiety 230 in 63% yield. Removal of the silicon and reduction on the double bond gave compound 223, (+)-Gigantecin in 69% yield.
37
Scheme 43: Silicon tethered total syntheses according to Hoye et al. (2016)
As illustrated in scheme 44, Quinn et al. introduced a methodology to synthesize building blocks for squamostanin C in a C2-symmetrical fashion: both THF moieties (x and y) were tethered by the present Ph2SiCl2 and coupled with a Grubbs metathesis in one pot.84
The synthesis started from diol 231, which underwent a desymmetrization where one hydroxyl group is replaced by an ester. Afterwards, dimerization was performed toward precursor 232 in 68% overall yield. Grubbs metathesis yielded in triene 233 in 83% yield. Subsequently, 3 steps were performed to transform triene 233 into diene 234 in 85% overall yield. Then, cyclization resulted in the desired core of squamostanin C in 51% by Hartung catalyst.85,86
38
Scheme 44: Triple Grubbs metathesis in an “one-pot reaction” toward the squamostanin C bis-THF core Besides this approach, Quinn reported another strategy as well in 2014, which is illustrated in Scheme 45.87 The second approach of Quinn et al. in 2014 is depicted in Scheme 45, that illustrated the synthesis toward the bis-THF core of cis-sylvaticin. This approach is known as the proof of concept of the approach illustrated in Scheme 44, since the triple metathesis is the most efficient pathway. The synthesis started from epoxide 236, which is treated by iPr2SiCl2 and 2,6-lutidine in dichloromethane to obtain bis-epoxide 237 in 74% yield. Subsequently, bis-epoxide 237 is transformed into diene 238 with four steps in 59% overall yield. Dihydroxylation and ring formation were performed in an overall yield of 55%, yielding the desired THF core of cis-sylvaticin 239. Scheme 44 illustrated an improved strategy, due to triple cross metathesis was performed to obtain the both THF moieties and the attachment of the building blocks in one step.
Scheme 45: Single Grubbs metathesis toward the cis-sylvaticin bis-THF core
The final approach was reported by Brown et al. in 2008 (Scheme 46).88 The first building block (243) was prepared from alkyne 240. Alkyne 240 was converted to precursor 241 (four steps in 67% overall yield), that subsequently underwent a cyclization reaction in 67% yield, obtaining THF moiety 242. Afterwards, building block 243 was prepared in five steps from 242 in 64% overall yield. Furthermore, the other building block (247) was obtained from bromide 244, which was converted to diene 245 using eight steps in 31% overall yield. Ring formation was performed toward THF-moiety 246 in 56% yield, which was transformed into the desired building block 247 by five steps in 41% yield. Subsequently, both building blocks were attached together, yielding bis-THF moiety 248 in 59%. The attachment was performed in three steps. Firstly, installing the silicon fragment for the unimolecular ring closing cross metathesis. Secondly, the attachment via Grubbs cross metathesis. Finally, the double bonds were replaced by single
39
bonds using palladium on carbon hydrogenation. Afterwards, the total synthesis toward cis-sylvaticin (249) was completed by four steps in 64% overall yield.
40
Conclusion
In this literature study, a comprehensive overview of the synthetic approaches toward bis-THF cores known in literature was shown in which the synthetic approaches were discussed.
The syntheses toward Annonaceous Acetogenins consist an interesting application due to its potency for anti-cancer treatments.1 The reason to investigate the use of these natural products rather than other new therapies against cancer is the interesting mechanism of action. Acetogenins are able to stop the energy production of cells, which is a novel approach of anti-cancer treatments. In this manner, the medicinal application of acetogenins toward tumor metastasis could become expanded on places that is challenging for traditional treatments. Moreover, there is an experimental challenge to pursue through novel synthesis methodologies via enzymatic reactions.
Many procedures were discussed for the synthesis of bis-THF cores. In the non-adjacent bis-THF core synthesis, an elegant approach has been depicted in Scheme 44. Three regioselective Grubbs cross metathesis reactions in one-pot have synthesized the desired THF core 235 in only 7 steps, where removal of the protecting groups must require an additional step. The obtained product is a C2-symmetrical building block, even as many examples in the adjacent bis-THF core synthesis.
The ideal pathway is hard to conclude. For example, Scheme 18 illustrates an efficient method for the step by step installation of THF moieties in the final THF core. This approach might be useful in the synthesis of probes for studies to different THF cores, using a general procedure. Furthermore, total syntheses were reported using no desymmetrization steps, however the synthesis toward C2-symmetrical building blocks were obtained in less steps. The unpublished approach of van Lint/Braun et al. illustrates the bio-inspired pathway to the synthesis of bullatacin. However, both this approach and the novel approach of van Lint et al. (2019), describes the total syntheses to certain cores as “threo-trans-threo-trans-erythro” cores. To expand the possible synthesis toward cis-THF cores, metal catalyzed reactions are useful methodologies. In addition, the use of epoxides might be useful for the synthesis toward rare cores. Moreover, the epoxide must be installed in an enantioselective manner. In contrast of the approach depicted in Scheme 33, both iodoetherification and the use of SN2 strategies resulted in enantiospecific reactions.
In order to obtain the highest overall yields, the most efficient way to synthesize adjacent bis-THF cored Annonaceous Acetogenins must be obtained via C2-symmetical building blocks, instead of using a linear approach toward the desired natural product. Efficient desymmetrization steps are not known yet, although metal catalyzed approaches are useful as well to avoid these challenging steps.
Research toward epoxide hydrolases would be required to synthesize the entire collection of adjacent bis-THF cores in the bio-inspired methodology of the unpublished results by van Lint/Braun et al.
41
Recommendations
In addition to the discussed approaches in the thesis, more work has been published. However, these approaches did not cover enough research for a good investigation/comparison. For example, Kumar et al. (2002) published an article about single electron oxidation methodology toward the adjacent bis-THF core.89 Since the approaches of single electron oxidations became “hot-topic”, it might be useful to investigate the strategy better. Moreover, the step by step installation that was depicted in Scheme 18, has been applied to the synthesis of bis-THF cores in Scheme 39. Besides that, this approach is used in the synthesis toward adjacent bis-THF cores. However, in the best of our knowledge, only one publication is reported about this approach.77 Further investigation into this topic might be useful for studies toward probes that have different core sizes.
Furthermore, since desymmetrization steps cannot be performed on an efficient manner, enzymatic approaches for desymmetrization could be useful. Moreover, enzymatic approaches are needed to develop more routes via cascade reactions toward C2-symmetrical building blocks.
Acknowledgements
I want to thank Romano Orru for having me in the group. I want to thank John Braun for supervising me during the literature study and the pleasant collaboration we had concerning the rush for finishing the study before I went to the United States. I also enjoyed that I was able to join some activities in the research group, such as soccer, Friday afternoon drinks, a symposium, and the facilities of having a workspace in the O2 building. I also participated in group meetings although I did not do work in the lab. I think there was a great participation of all members of the SyBOrCH group in order to allow me in the research group, even the period of the literature study is only two months without experimental work. I really appreciated the atmosphere of the colleagues in this short time.
