Remote Stereocontrol in Aryl/Alkyl C–H Insertion Reactions of Rhodium Carbenes: Assembly of Pseudorigidol A and B

Abstract

The first total syntheses of pseudorigidols A and B are reported. Our synthesis employs a rhodium-catalyzed C–H insertion reaction to build the tricyclic core. The catalyst screen revealed the influence a remote stereogenic center has on the diastereoselectivity of the insertion. Using two different catalysts, we were able to synthesize the two natural products with high diastereoselectivity. These results will lay the foundation for further studies on the influence of remote stereogenic centers.

Graphical Abstract

Sesquiterpene natural products have become popular total synthesis targets due to their challenging ring systems and biological activity.^1-5 When considering the different families of sesquiterpenes synthesized, there have not been many reported syntheses of cadinane sesquiterpenes.^6-12 Furthermore, bicyclic cadinanes are common, but tricyclic cadinanes are quite rare and not many have been isolated. Pseudorigidol A (1) and pseudorigidol B (2) were recently isolated as a diastereomeric pair from the Caribbean sea plume Pseudopterogorgia rigida (Figure 1).¹³ The relative stereochemistry of these cadinane-type sesquiterpenoids was assigned through 1D NOE experiments while the absolute configuration was unassigned. Although the alkyl stereogenic centers would be difficult to access by traditional methods, we recognized that insertion of an aryl/alkyl rhodium carbene could be an effective strategy for these targets.^14,15 While this approach was effective for the related core of cycloartobiloxanthone¹⁶, the influence remote stereocenters have on intramolecular C–H insertion reactions has been underexplored.^17-23 We envisioned two possible strategies that to control the relative configuration of C7 and C10 to access the two different targets from a common substrate. If the methyl group at C7 exhibited little influence on the C–H insertion reaction at C10, then racemic precursor 3 could yield enantiomerically pure diastereomers 1 and ent-2 with different configurations at C7 (catalyst control, Figure 1). Alternatively, if the C7 methyl group influenced the outcome of the C–H insertion at C10, then enantiomerically pure substrate 4 could lead to 1 and 2 (with the same configuration at C7) using different catalysts. Extensive exploration of both strategies revealed interesting levels of influence from the remote stereogenic center and ultimately the latter route was successful. Here we report the successful synthesis of pseudorigidols A and B, confirming the assignment of relative stereochemistry and establishing their absolute configurations.

Figure 1.

Two approaches to access pseudorigidol A and B.

The synthesis of (+/−)-3 began with a Friedel-Crafts acylation of 7 using methyl 4-chloro-4-oxobutanoate (Scheme 1). The benzylic ketone 9 was olefinated with methylenetriphenylphosphorane and hydrogenated to yield (+/−)-10. This intermediate was then saponified and carried into an intramolecular Friedel-Crafts to afford tetralone core (+/−)-11. Regioselective cleavage of the methyl group ortho to the ketone was achieved using BCl3. The resulting phenol was alkylated with isopropyl iodide and treated with hydrazine and acetic acid at elevated temperature to afford hydrazone (+/−)-3.

Scheme 1.

Synthesis of Hydrazone (+/−)-3

Oxidation of hydrazone (+/−)-3 to the diazo species with subsequent C–H insertion was carried out utilizing a one-pot sequential process (Table 1). To identify the catalysts that would allow the natural products to be accessed selectively, a screen using dirhodium catalysts with varying ligands was conducted. The objective of the screen was to assess the extent to which the stereochemical outcome of the reaction was influenced by either catalyst or substrate control, which includes the influence the remote stereocenter may have. If the substrate contributes highly to the outcome of this insertion, the diastereomeric ratio (dr) would be high (>95:5) regardless of the catalyst used. In the case of catalyst control, the dr would be more strongly dictated by the steric and electronic environment of each catalyst. For example, when employing an achiral catalyst, if the dr is around 50:50, it would indicate that the substrate itself has little impact on the diastereoselectivity on the reaction. Conversely, if, when using an achiral catalyst, a stereochemical preference is preferred, it can be concluded that the diastereoselectivty is being influenced by the steric and electronic qualities of the substrate. With a bulky achiral catalyst (Rh₂(PTCC)₄), high selectivity for a trans relationship between the hydrogen on C10 and the C7 methyl group was observed. Moderate selectivity for a cis relationship between C10 and C7 was observed when a less bulky achiral catalyst (Rh₂(Mes)₄) was used. This indicates that the substrate indeed impacts the stereochemical outcome of the insertion. When Rh₂(S-BTPCP)₄ was used, the cisE1:transE1 dr was further enhanced, indicating cooperative influences from the substrate and catalyst. However, the dr of cisE2:transE2 was diminished. This result was a good indication of a match-mismatch effect between the substrate and the catalyst. Upon switching to the enantiomer of catalyst, the dr of cisE2:transE2 is enhanced, and as expected, the dr of cisE1:transE1 gets diminished. For accessing the cis diastereomer in high selectivity, Rh₂(S-TFPTTL)₄ gave an enhanced dr for cisE2:transE2, and diminished dr for cisE1:transE1. Based off the BTPCP ligand results, we inferred that Rh₂(R-TFPTTL)₄ will give an enhanced dr for cisE1:transE1. From these results, we can expect that 1 and 2 can be accessed selectively from using Rh₂(S-BTPCP)₄ and Rh₂(R-TFPTTL)₄, respectively.

Table 1.

Insertion Results from Hydrazone (+/−)-3

The last step of the route is to deprotect the methyl ether protected phenol ortho to the C7 methyl group (Scheme 2). Multiple attempts to deprotect the phenol were unsuccessful, giving unreacted starting material, or minor cleavage along with an inseparable and unidentifiable byproduct. Although cleavage of phenolic methyl ethers with BBr3 is well-established, the presence of ortho alkyl substituents drastically diminishes the viability of this method.

Scheme 2.

Unsuccessful Deprotection of (+/−)-12 and (+/−)-13

A similar route to 1 and 2 was developed using the more easily cleaved isopropyl protecting group.¹⁶ Racemic hydrazone 4a underwent one-pot oxidation and insertion with selectivities similar to (+/−)-3 (see SI). After confirming successful cleavage of the phenolic i-Pr group using AlCl₃, enantiomerically pure hydrazone 4a was prepared in a high-yielding sequence from 9 (Scheme 3). To access the desired C7 configuration, the hydrogenation from the racemic route can be replaced with an asymmetric reduction using (R,R)-[COD]Ir[cy₂PThrePHOX].^24,25 The protected precursors to 1 and 2 were accessed by using either Rh₂(S-BTPCP)₄ and Rh₂(R-TFPTTL)₄. Upon cleaving the isopropyl ether on 16 using AlCl₃, the ¹H NMR data revealed that the C10 center formed from the insertion had epimerized. Prior to deprotection, the dr of 16 was 3:97 (15:16), but after the deprotection, the dr was 62:38 (15:16). After re-evaluating the ¹H NMR data for the racemic material, we confirmed that epimerization had also occurred. We hypothesized that the benzodihydrofuran ring opened upon addition of Lewis acid and rearranged to yield benzylic alkene intermediate 18, lacking the C10 stereogenic center. Protonation of this alkene with little facial selectivity then leads to formation of a mixture of 1 and 2. The epimerization seemed to happen competitively with isopropyl group removal given that 1 and 2 were the only observed products.

Scheme 3.

Synthesis and Insertion Reaction of Hydrazone 4a

Two more protecting groups were evaluated, each cleavable with conditions other than a Lewis acid. The triisopropylsilyl (TIPS) group was chosen for its robustness (Scheme 4). Although the TIPS group on 21 did not survive saponification with NaOH conditions, cleavage of the methyl ester proceeded smoothly when TMSOK was used. The yield of 22 increased significantly when HFIP cyclization conditions reported by the Aubé group were used.²⁶ Unfortunately, when hydrazone 4b was subjected to one-pot oxidation and insertion, the diastereoselectivity was markedly lower than the methyl and isopropyl ether substrates. Aside from the low diastereoselectivity, ¹H NMR data revealed that 1,2-hydride shift of the Rh-carbene intermediate also occurred (see SI). Given the undesirable stereoselectivity and reactivity from the TIPS group, our efforts shifted to the second protecting group evaluated.

Scheme 4.

Synthesis and Insertion Reaction of Hydrazone 4b

The final protecting group investigated for the synthesis of 1 and 2 was the benzyl ether (Scheme 5). We reasoned that the steric demand would be similar to the isopropyl ether, thus avoiding the negative impact of the bulky TIPS ether. TIPS ether 23 was deprotected with TBAF, alkylated with benzyl bromide and condensed with hydrazine to yield substrate 4c. One-pot oxidation and insertion of 4c occurred with high diastereoselectivity and opposite senses of induction with S-BTPCP and R-TFPTTL, aligning closely with the results observed for both the methyl and isopropyl ether substrates.

Scheme 5.

Synthesis and Insertion Reaction of Hydrazone 4c

Hydrogenation of 27 and 28 under standard conditions led to 1 and 2, respectively, in high yields (Scheme 6). The ¹H and ¹³C NMR data of the synthetic materials matched the reported literature values. The sign of the optical rotation for 1 was consistent with the reported value (${[\alpha ]}_D^{23}$: −5.8 for synthetic material vs ${[\alpha ]}_D^{20}$: −30.0 for natural material). The optical rotation value for 2 was opposite in sign to the reported value (${[\alpha ]}_D^{23}$: +34.2 for synthetic material vs ${[\alpha ]}_D^{20}$: −16.9 for natural material). After discussion with the author of the isolation paper, discrepancies in magnitude may be attributed to impurities present during the isolation of the natural material. Compound 1 was acylated with nitrobenzoyl chloride, resulting in crystalline derivative 29. Single crystal X-ray diffraction enabled assignment of the absolute stereochemistry as R configuration at C7 and S configuration at C10 for pseudorigidol A. Therefore, the absolute stereochemistry for pseudorigidol B can be assigned as R configuration at both C7 and C10. Our absolute stereochemistry assignments are consistent with the reported relative stereochemistry assignments. However, what remains abstruse is the optical rotation of 2 being opposite in sign to the reported value. From our results, we believe that the isolated natural products are epimeric at C7 as opposed to C10, which was originally reported.

Scheme 6.

Synthesis and Stereochemical Assignment of 1 and 2

Therefore, what was isolated was 1 and ent-2, not 1 and 2.

In conclusion, we have completed the first syntheses of pseudorigidol A (1) and pseudorigidol B (2). The synthesis relies on the diastereoselective C–H insertion to form the final tricyclic core. The stereochemical outcome of the insertion is influenced by the remote stereocenter at C7, along with the steric bulk of the protecting group on the phenol ortho to C7.

Supplementary Material

¹H, ¹³C NMR spectra, experimental procedures, X-ray crystallography data, and chiral HPLC spectra.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.orglett.6c00438?goto=supporting-info.

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.