Convergent Total Syntheses of (−)‐Rubriflordilactone B and (−)‐pseudo‐Rubriflordilactone B

Abstract A highly convergent strategy for the synthesis of the natural product (−)‐rubriflordilactone B, and the proposed structure of (−)‐pseudo‐rubriflordilactone B, is described. Late stage coupling of diynes containing the respective natural product FG rings with a common AB ring aldehyde precedes rhodium‐catalyzed [2+2+2] alkyne cyclotrimerization to form the natural product skeleton, with the syntheses completed in just one further operation. This work resolves the uncertainty surrounding the identity of pseudo‐rubriflordilactone B and provides a robust platform for further synthetic and biological investigations.

To a stirred suspension of (iodomethyl)triphenylphosphonium iodide (2.22 g, 4.18 mmol, 1.8 equiv.) in a flame dried flask in THF (35 mL) under argon at RT was added NaHMDS (1.98 mL,3.95 mmol,2 M in THF,1.7 equiv.). After 20 min, the solution was cooled to -78 °C and a solution of the crude product from the previous step in THF (15 mL) was added. After 15 min, the reaction mixture was warmed to rt over 30 min until TLC showed complete formation of the cis-vinyl iodide derivative.
The reaction mixture was then cooled to -78 °C, and additional NaHMDS (3.50 mL, 7.00 mmol, 2 M in THF, 3.0 equiv.) was added. The reaction mixture was warmed to rt over 30 min and stirred for a further 10 min.
The reaction was monitored by TLC and on completion, 0.5 mL MeOH was added. The reaction was stirred for 2 h followed by separation of the layers. The aqueous layer was extracted with CH2Cl2 (3 x 25 mL) and the O O Br Br S9 combined organic layers were dried (Na2SO4) and concentrated. The crude material was carried forward to the next step without further purification.
To a solution of PPh3 (713 mg, 2.56 mmol, 5.0 equiv.) in CH2Cl2 (5 mL) at 0 °C was added CBr4 (425 mg, 1.28 mmol, 2.5 equiv.). The mixture was stirred at 0 °C for 15 min before being cooled to -30 °C. To this was added dropwise a solution of the crude aldehyde and Et3N (0.71 mL, 5.13 mmol, 10 equiv.) in CH2Cl2 (3 mL), and the reaction mixture was warmed to 0 °C over 1h. The reaction was quenched by addition of with saturated aq. NH4Cl, the layers were separated and the aqueous layer extracted with DCM (3 x 15 mL). The combined organic layers were dried (Na2SO4) and concentrated, and the residue was purified via flash column chromatography
After separation of the layers, the aqueous phase was extracted with diethyl ether (3 x 200 mL). The organic layers were combined, dried (MgSO4), filtered, and concentrated. Flash chromatography of the residue provided the β-ketoester S7 (4.30 g, 18.54 mmol, 75%) as a slightly yellow oil (keto-enol mixture).
To a stirred solution of crude triol in CH2Cl2 (180 mL) was added NaIO4 on silica (10 wt%, 51.9 g, 24.4 mmol, 3.0 equiv.). The suspension was stirred for 12 h before being filtered through a short plug of silica. Following concentration, the crude lactol was carried forward without further purification.
The residue was used directly in the next step without further purification.
Et2O was added, and the aqueous layer was extracted with diethyl ether (3 x 50 mL). The combined organic layers were dried with MgSO4, and concentrated. The crude product was purified by flash chromatography (19:1 petroleum ether / Et2O) to give S13 (193 mg, 0.551 mmol, 92%) as a light yellow oil.
To a solution of PPh3 (662 mg, 2.53 mmol, 5.0 equiv.) in CH2Cl2 (7 mL) at 0 °C was added CBr4 (419 mg, 1.26 mmol, 2.5 equiv.). The mixture was stirred at 0 °C for 15 min, then cooled to -30 °C. To this, a solution of the crude aldehyde and Et3N (0.70 mL, 5.05 mmol, 10 equiv.) in CH2Cl2 (3 mL) was added dropwise, and the reaction mixture was allowed to warm to 0 °C over 1h. The reaction was quenched with saturated aq.
This mixture was transferred to a flame-dried flask as a solution in THF (10 mL). The solution was cooled to -78 °C, then LiHMDS (2.05 mL, 2.05 mmol, 1 M in THF, 4.0 equiv.) was added. The resulting solution was warmed slowly to -15 °C over 2 h, then cooled again to -78 °C. n-BuLi (0.61 mL, 1.51 mmol, 2.5 M in hexanes, 3.0 equiv.) was added dropwise along the walls of the vial, and stirred for 5 min. The reaction was monitored by the TLC and on completion, a solution of saturated aq. NH4Cl (10 mL) was added at -78 °C.

Stereochemical assignment of Diynes (8a-8d)
To assign the stereochemistry of the four diastereomers produced during the connection of the F and G rings, a combination of nOe enhancements and coupling constant analysis was used. In adduct 8b, the C22 stereocentre was assigned as S, with this proton on the bottom face of the ring (as drawn) due to strong nOe enhancements with the protons on the C21 methyl and at C17. The C23 stereocentre was assigned as R (proton on the top face as drawn) due to a combination of a strong enhancement of the C23 proton with the proton at C20, an enhancement between H24 and H22, and a coupling constant J(H22-H23) of 7.5 Hz. These observations would be explained by a dominant conformation similar to that in Figure S1b, in which the H23 and H22 protons are antiperiplanar. Figure S1. a. Strong and medium enhancements seen in the nOe spectrum for compound 8b; b. Conformation of butenolide ring in compound 8b.
The C22 stereocentre of adduct 8c was assigned as R due to the lack of corresponding strong nOe enhancements of H22 with the protons at C16 and C17 on the bottom face (see above, and discussion on 8a below).
Assignment of the C23 stereocentre as S (bottom face as drawn) was based on strong enhancements of H23 S25 with the C21 methyl protons in combination with the coupling constant for J(H22-H23) of 9.3 Hz, again indicating an antiperiplanar orientation. These observations would be explained by the conformation shown in Figure S2. The configuration of 8c was also confirmed by x-ray crystallographic analysis. (See P S48 for details) Figure S2. a. Strong enhancements seen in the nOe spectrum for compound 8c; b. Conformation of butenolide ring in compound 8c.
Compounds 8a and 8d were obtained as an inseparable mixture. After assignments of all peaks by 1 H-1 H COSY, cross peaks in the 1 H-1 H NOESY spectrum were used to provide stereochemical information on these final two diastereomers. The proton at C22 in one diastereomer displayed strong enhancements with H17 and the methyl group C21, indicating that it was oriented on the bottom face, which matches the proposed C22S stereochemistry of the natural product. This diastereomer was named 8a. The proton at C23 displayed a medium strength enhancement with the proton at C20 and the methyl group at C21. A coupling constant for H22-H23 of ~3 Hz, and the through-space enhancement seen between H22 and H24 in the butenolide ring, would all be explained by an S configuration at C23, and the conformation shown in Figure S3. The final diastereomer (8d) displayed a lack of strong enhancements between the H22 and either the H16 or H17 signals, which indicated an R configuration (when compared with other diastereomers), with the proton at C22 being on the top face of the F ring. The proton at C23 showed a strong enhancement with the methyl group at C21, which combined with a coupling value of 1.9 Hz, suggested the 23R stereochemistry in which the butenolide adopts a conformational rotation placing the H22 and H23 close to a 90° angle, as shown in Figure S4.   8S,9R,10aR)-1-Hydroxy-8-methyl-1,2,3,4,5,7b,8,9,10a,11-decahydrocyclohepta[4,5]indeno-[2,1b]furan-9-yl acetate, 28 To a solution of 27 (15.0 mg, 0.039 mmol, 1.0 equiv.) in THF (1.0 mL) at rt was added AcOH (2.3 µL, 0.039 mmol, 1.0 equiv.), and then TBAF (78 µL, 0.078 mmol, 1 M in THF, 2.0 equiv.). The reaction was stirred for 15 min, after which time starting material had been consumed. The reaction was then quenched by addition of saturated aq. NH4Cl (10 mL). The aqueous layer was extracted using EtOAc (4 x 10 mL). The combined organic layers were dried using Na2SO4, and concentrated. The crude product was used directly in the next step without further purification.
To a solution of 2-nitrophenyl selenocyanate (15 mg, 0.063 mmol, 3.0 equiv.) in THF (0.2 mL) under inert atmosphere at rt was added n-Bu3P (13 mg, 0.063, 3.0 equiv.). The solution was stirred for 15 min following which a solution of the crude in THF (0.5 mL) from the previous step was added. The reaction was stirred for 2 h at rt before quenching with the NH4Cl (sat. sol.). The mixture was poured into EtOAc and the layers were separated. The aqueous layer was extracted with EtOAc and the organic layers were combined, washed with water, brine, and then dried over Na2SO4. The solvent was removed to give a residue which was passed through a short plug of silica to remove excess of 2-nitrophenyl selenocyanate and n-Bu3P. The resulting mixture was used for the next step without further purification.
To the above crude, 1 mL THF was added followed by H2O2 (50 wt% solution in water, 15µL, 0.21 mmol, 10.0 equiv.) at rt. The resulting solution was stirred for 15 min before being quenched with Na2SO3 (sat. sol.).
After stirring for 15 min, the mixture was poured into EtOAc (10 mL) and the layers were separated. The aqueous layer was extracted with EtOAc (4 × 10 mL) and the organic layers were combined, washed with water, and then dried over Na2SO4. The solvent was removed to give a residue which was purified using column chromatography to afford alkene 32 (3.9 mg, 0.008 mmol, 40% over all yield).

Acetate 33
To a stirred solution of RhCl(PPh3)3 (2.0 mg, 0.003 mmol, 0.1 equiv.) in anhyd. DCE (2 mL), a solution of 31 (10 mg, 0.021 mmol, 1.0 equiv.) in DCE (3 mL) was added. The reaction was stirred at 50 °C. After 12 h, on completion of the reaction, the reaction vessel was cooled to rt and DCE was evaporated. The residue was dried under vacuum.
To a flame dried vial, in the glovebox, Martin's sulfurane was added (25 mg, 0.031 mmol, 1.5 equiv.). To it, 1 mL dry DCM was added followed by a solution of the crude residue (in 1 mL DCM) from the previous step.
The reaction was stirred for 10 min before quenching with 5 mL NaHCO3 solution. The aqueous layer was extracted with EtOAc (4 × 10 mL) and the organic layers were combined, washed with water, and then dried over Na2SO4. The solvent was removed to give a residue which was purified using column chromatography to afford acetate 33 (10 mg, 0.015 mmol, 70% over two steps).

Data comparison tables of natural products
Comparison of spectroscopic data for rubriflordilactone B:

S50
Discussion of proposed misassignment at H11/C11, H12/C12, and C15: There are significant discrepancies between the spectroscopic data for synthetic pseudo-rubriflordilactone B, and the isolation paper.
C15 (original assignment 30.7 ppm): The inconsistency of this carbon resonance was noted by Sarotti and Kaufman (OL 2016, 6420), who identifed HMQC correlations of this carbon with protons around 1.2 ppm in the isolation spectrum. This revealed a misassignment of this peak.
We found that the C15 peak was very difficult to identify, as its resonance is extremely broad at ambient temperature, and can barely be resolved from the baseline. By acquiring the 13 C NMR spectrum at 60 °C, a broad resonance was revealed at 41.7 ppm, which we assign as C15. Notably, this revised assignment is much more consistent with C15 in rubriflordilactone B (39.5 ppm).

H11,C11/H12,C12:
We believe that the peaks originally assigned as H11 and H12 (7.53 and 7.89 ppm) in fact correspond to an isolation artifact, most likely phthalate impurities which commonly arise from plastic labware. Comparison of the 1 H NMR spectrum for dibutyl phthalate in d5-pyridine ( Figure S5a) with the isolation paper reveals a compelling match between the aromatic signals at 7.5 and 7.9 ppm ( Figure S5b).
Similarly, The dibutyl phthalate 13 C NMR spectrum shows C-H carbons at 129.7 and 131.8 ppm ( Figure   S6a). These match with carbons at 129.3 and 131.5 ppm in the isolation paper (expansion in Figure S6b), which show HMQC cross-peaks (in the isolation paper) with the protons (Expansion Figure S6c). A further phthalate resonance at 133.5 ppm could correspond to a resonance in the isolation paper at 133.2 ppm. The following peaks are affected; resonances that are broad or emerge only at 60 °C are listed in blue, and signals that (in our work) are not resolved at 60 °C despite long acquisition times are shown in red. Notably, these signals are clustered around the 7-membered C-ring, where it is reasonable to expect conformational interconversion on the NMR timescale. An consequence is that the corresponding cross-peaks in 2D NMR spectra are also not observed, which hinders complete identification of all carbon signals in the molecule.

X-ray Crystallography
Low temperature single crystal X-ray diffraction data were collected for 1, 7b and 16a with a (Rigaku) Oxford Diffraction SuperNova A diffractometer at 150 K, and data for 8c were collected using I19-1 at the Diamond Light Source [1] at 100 K. All data were reduced using CrysAlisPro, solved using SuperFlip [2] and the structures were refined using CRYSTALS. [3] Structure 7b contained solvent accessible voids comprising of weak, diffuse electron density. The discrete Fourier transforms of the void regions were treated as contributions to the A and B parts of the calculated structure factors using PLATON/SQUEEZE [4] integrated within the CRYSTALS software. This enabled a comparison of models, one of which contained the disordered solvent, the other without. The change in R index (all data) was small, 3.10 versus 3.06 without the disordered solvent, and the model without the disordered solvent was thus chosen.
The Flack x parameter [5] was refined in all cases. Bayesian analysis of the Bijvoet pairs was also carried out using all the data used in the refinement. [6] This gave the Hooft y parameter, the P2 probability (the likelihood that the hand is correct given the crystal was enantiopure), and the P3 probability (the likelihood that the hand is correct given the crystal was enantiopure or racemic).
Further details about the refinements, including disorder modelling and restraints, are documented in the CIF.