Sustainable Syntheses of (−)-Jerantinines A & E and Structural Characterisation of the Jerantinine-Tubulin Complex at the Colchicine Binding Site

The jerantinine family of Aspidosperma indole alkaloids from Tabernaemontana corymbosa are potent microtubule-targeting agents with broad spectrum anticancer activity. The natural supply of these precious metabolites has been significantly disrupted due to the inclusion of T. corymbosa on the endangered list of threatened species by the International Union for Conservation of Nature. This report describes the asymmetric syntheses of (−)-jerantinines A and E from sustainably sourced (−)-tabersonine, using a straight-forward and robust biomimetic approach. Biological investigations of synthetic (−)-jerantinine A, along with molecular modelling and X-ray crystallography studies of the tubulin—(−)-jerantinine B acetate complex, advocate an anticancer mode of action of the jerantinines operating via microtubule disruption resulting from binding at the colchicine site. This work lays the foundation for accessing useful quantities of enantiomerically pure jerantinine alkaloids for future development.

Melting point data were collected using a Stuart SMP3. Optical rotations were recorded using ADP440 polarimeter.
Voacanga africana seeds (200 g) were ground into a fine powder and suspended in an aqueous solution of 1% H2SO4 (2.0 L). The resulting mixture was then stirred at room temperature for 16 h. The acidic extract was then filtered and NaCl (200 g) added to the filtrate, and stirred for 16 h. CHCl3 (1.5 L) was then added and the biphasic solution vigorously stirred for 3 h. After settling, the aqueous layer was removed and discarded and the remaining organic emulsion filtered through Celite ® . The resulting solution was then separated and the organic fraction dried over anhydrous MgSO4 and filtered. The solvent was then removed under reduced pressure and the resulting material dissolved in CHCl3 (50 mL), followed by the addition of 30% aqueous ammonia (30 mL

Scheme S-3. Boc protection of (-)-15-iodo-tabersonine.
To a stirred solution of (-)-15-iodo-tabersonine (11.9 g, 25.7 mmol) in anhydrous THF (250 mL) was added Boc2O (8.20 mL, 77.1 mmol) and DMAP (313 mg, 2.57 mmol). The resulting mixture was heated at 60 °C for 18 h. The solvent was then removed under reduced pressure and the residue dissolved in CH2Cl2 (200 mL), followed by washing with water (3 x 50 mL). The organic fraction was dried over anhydrous MgSO4 and filtered. The solvent was removed under reduced pressure to yield the crude product, which was

Agent stocks
Jerantinine A (1), jerantinine A acetate (7) and colchicine were provided as solids and reconstituted with DMSO to yield concentrations of 10 mM. Stocks were stored as 10 µL aliquots at −80 °C protected from light.

MTT assay
The 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide (MTT) assay, adapted from Mosmann 4 , was used to assess the ability of test agents to inhibit cell growth and/or evoke cytotoxicity 5 . MTT assays were performed at the time of agent addition (T zero) and following 72 h exposure of cells to test agents, as previously described 6 .

Clonogenic assay
The clonogenic cell survival test measures the ability of a single cell to survive brief exposure to test agents and maintain proliferative potential to form progeny colonies 7,8 .
The assay was performed as previously described 6 .

Cell cycle analysis
Cell cycle analysis was carried according to Nicoletti et al. 9 Cells were seeded in cell culture dishes at densities of 3-5 × 10 5 cells/dish in 10 mL medium. Following treatment, cells were harvested and pelleted by centrifugation then re-suspended in 0.5-1 mL fluorochrome solution (50 µg/mL propidium iodide (PI), 0.1 mg/ mL ribonuclease A, 0.1% v/v Triton X-100, and 0.1% w/v sodium citrate in dH2O). Cells were stored overnight in the dark at 4 °C. Cell cycle analyses were performed on a Beckman Coulter FC500 flow cytometer. EX-PO32 software was used to analyse data.

Confocal Microscopy
Confocal imaging was performed as previously described. 13 Procedures were performed at room temperature. Cells were fixed in formaldehyde (3.7% in PBS; 10-15 min) then permeabilised by PBT (PBS + 0.1% Triton X-100; 2-3 min). Blocking agent (PBT + 1% BSA; 1 h) was used to prevent binding non-specific protein binding. Cells were incubated with 1° Ab (monoclonal anti α-tubulin Ab, VWR International Ltd.; 2 h), washed with PBT before incubation in the dark with 2° Ab for 1 h. Cells were incubated with DNA binding dye (DRAQ5) for 5 min in the dark; a Zeiss LSM510 Meta confocal microscope was used to capture images. Quantification was completed using ImageJ, calculating the area for each colony. Multiple comparisons following one-way ANOVA of 96 h images post agent addition were calculated with GraphPad Prism v7.0.  20,000 events were analysed per sample. Jerantinine A and jerantinine A acetate caused multinucleation (1), nuclear fragmentation (2) and multipolar spindle formation (3). (2) (1) (2) (3)

Confocal Microscopy
After 24 h of treatment, extensive mitotic disruption was observed. Multinucleation, nuclear fragmentation and multipolar spindles were clearly evident in treated cells exclusively; the morphology of non-treated cells demonstrated normal cell division.

Discussion
Jerantinine A (1) and jerantinine A acetate (7) are aspidosperma indole alkaloids, and have shown profound growth inhibitory and cytotoxic activity against human-derived MCF-7 and HCT-116 cancer cell lines. Jerantinine A acetate (7) had slightly higher potency, and would represent a better drug candidate due to its increased stability. Additionally, the presence of an acetate group reduces overall polarity and could help intracellular access across hydrophobic cell membranes.
Confocal microscopy, following treatment of cells with synthetic jerantinine A (1) and jerantinine A acetate (7)

Crystallisation, data collection and structure determination
Crystals of T2R-TTL were generated as described previously 15,16

Methods
The X-ray crystal structures of the colchicine-tubulin complex (PDB ID 4o2b) 24 , the vinblastine-tubulin complex (PDB ID 5j2t) 25 , the structure of the taxol-tubulin complex determined by electron crystallography (PDB ID 1jff) 26 , and the structure of the (-)jerantinine B acetate-tubulin complex determined here, were used in in silico docking studies. Comparative modelling was performed on the tubulin structures to introduce the human tubulin amino acid sequence using the MODELLER software 27 . Ligand structures were minimised at the B3LYP/6-31G(d) level using the GAUSSIAN-03 software 28 . Docking was performed using the Vina software, 29 implemented in YASARA (www.yasara.com).
The highest-scoring docking solutions from the Vina analysis were subjected to local geometry optimisation; the binding energies from these calculations are reported in Table   2.

Results
The structure of both a-and b-tubulin comprises six alpha helices (H1-H6), six beta strands (S1-S7), and six loops (T1-T6) connecting alternating helices and strands 26 . Taxol binds only b-tubulin, interacting with residues in helix H1, the H6-H7 loop, the H7 helix, the S7-H9 loop, and the M-loop -this is often referred to as the taxane site; binding to this site results in microtubule stabilisation. On a-tubulin, vinblastine interacts with residues in H10, S9 and the T7-loop, and on b-tubulin with residues in the T5-loop and H6-H7 loop. Ligands binding to the vinblastine site also destabilise microtubules.
Jerantinine B acetate binds the colchicine binding pocket, also at the interface between aand b-tubulin subunits; this pocket is formed by residues in the T5-loop on a-tubulin, and H7, H8, S8, S9 and the T7-loop on b-tubulin. Thus, the three sites, taxane-, colchicineand vinblastine-, are distinct and non-overlapping 30 .