Hydroxylated Rotenoids Selectively Inhibit the Proliferation of Prostate Cancer Cells

Abstract

Prostate cancer is one of the leading causes of cancer-related death in men. The identification of new therapeutics to selectively target prostate cancer cells is therefore vital. Recently, the rotenoids rotenone (1) and deguelin (2) were reported to selectively kill prostate cancer cells and the inhibition of mitochondrial complex I was established as essential to their mechanism of action. However, these hydrophobic rotenoids readily cross the blood brain barrier and induce symptoms characteristic of Parkinson’s disease in animals. Since hydroxylated derivatives of 1 and 2 are more hydrophilic and less likely to readily cross the blood brain barrier, 29 natural and unnatural hydroxylated derivatives of 1 and 2 were synthesized for evaluation. The inhibitory potency (IC₅₀) of each derivative against complex I was measured and its hydrophobicity (Slog₁₀P) predicted. Amorphigenin (3), dalpanol (4), dihydroamorphigenin (5), and amorphigenol (6) were selected and evaluated in cell-based assays using C4-2 and C4-2B prostate cancer cells alongside control PNT2 prostate cells. These rotenoids inhibit complex I in cells, decrease oxygen consumption, and selectively inhibit the proliferation of prostate cancer cells, leaving control cells unaffected. The greatest selectivity and anti-proliferative effects were observed with 3 and 5. The data highlight these molecules as promising therapeutic candidates for further evaluation in prostate cancer models.

Prostate cancer is one of the leading causes of cancer-related death in men.^1,2 Prostate tumors generally respond well to androgen deprivation therapy and shrink, but then almost invariably become androgen insensitive, regrow autonomously, and metastasize, often to bone.^3,4 The identification of new therapeutic agents to selectively target prostate cancer cells is therefore vital. Recently, the natural rotenoids rotenone (1) and especially deguelin (2) (Figure 1) were shown to selectively kill phosphatase and tensin homologue deleted on chromosome 10 (PTEN)-null prostate cancer cells.⁵ The inhibition of mitochondrial complex I (NADH:ubiquinone oxidoreductase) was established as essential to their mechanism of action and selective toxicity.⁵ This finding further highlights the therapeutic potential of 1 and 2 ^6–11 and is consistent with their being canonical ubiquinone-binding site inhibitors of complex I.^12–15 Complex I is a major entry point for electrons into the mitochondrial respiratory chain and an integral contributor to oxidative phosphorylation (OxPhos)¹⁶ that is a target for other potential anticancer compounds, such as metformin^17,18 and IACS-010759¹⁹ from the MD Anderson Cancer Center. It oxidizes the NADH produced by the citric acid cycle and other metabolic pathways, transfers the electrons to ubiquinone to sustain the reduction of molecular oxygen through complexes III and IV, and transports protons across the mitochondrial inner membrane to maintain the proton motive force that drives ATP synthesis and metabolite transport.¹⁶ In cells, 1 and 2 inhibit complex I and disrupt OxPhos, leading to a decrease in mitochondrial ATP production, hydrolysis of ATP to maintain the mitochondrial membrane potential, and ultimately cell death by apoptosis.⁵ Selective toxicity towards prostate cancer cells arises because these cells require greater than normal amounts of ATP to sustain unchecked growth, following loss of the tumor suppressor PTEN early in tumorigenesis, and this ATP is produced most efficiently by OxPhos.⁵ PTEN-null prostate cancer cells are therefore highly dependent on OxPhos and especially vulnerable to inhibitors of complex I, while PTEN-positive normal prostate cells remain relatively tolerant of them.⁵

Figure 1.

The structures of rotenoids 1-8 with ring-labels and atom numbers. Molecules marked * are both plant secondary metabolites and products of detoxifying metabolism in mammals, fish, and insects.^31,32

While rotenone (1) and particularly deguelin (2) thus appear as promising therapeutic agents, their neurotoxicity must be properly considered. Both 1 and 2 cross the blood-brain barrier (BBB) in rats and mice, cause neuronal damage when administered in doses of 3 and 6 mg kg^-1 day^-1, respectively, and induce symptoms characteristic of Parkinson’s disease.²⁰ It is clear that 2 is less neurotoxic than 1 because it is more susceptible to oxidation and so metabolized and cleared more rapidly,²⁰ but it is not clear whether 2 is suitable for long-term systemic administration.²¹ Importantly, recent work⁵ suggests that rotenoids related to 1 and 2 that inhibit complex I should also selectively kill prostate cancer cells. The pool of potential rotenoid therapeutic agents can therefore be expanded to include those with potentially lower neurotoxicity profiles.

To a first approximation, rotenone (1) and deguelin (2) readily cross the BBB because they are hydrophobic.^22,23 Accordingly, hydroxylated rotenoids such as amorphigenin (3),^24,25 dalpanol (4),^26,27 dihydroamorphigenin (5),²⁸ and amorphigenol (6) (Figure 1),^29–32 which by comparison are more hydrophilic than 1 and 2, would not be expected to readily cross the BBB. Hydroxylated metabolites of 1 and 2 were analyzed for in the rat model of Parkinson’s disease, but only rotenolone (7) and tephrosin (8) (Figure 1) were detected in the brains of treated animals.²⁰ Hydroxylated rotenoids that are more hydrophilic than 7 and 8, such as 3 and 6, could thus constitute potentially promising therapeutic agents for the treatment of prostate cancer.

However, although hydrophobicity is an important factor in determining whether a molecule can cross the BBB,^22,23 the model outlined above is simplistic. Many additional factors can influence whether a molecule can cross the BBB, including molecular weight, volume, topological molecular polar surface area, solvent accessible surface area, and the number of hydrogen bond donors and acceptors.^33,34 Molecules can also be actively expelled from the brain by various efflux transporters, such as P-glycoprotein, and their accumulation thereby prevented.³⁴ Moreover, mice administered with 100-900 mg kg^-1 of plant extracts rich in dalpanol (4) were reported to suffer severe neurological impairments,³⁵ although histopathological analysis of the brain tissue of treated mice did not show any differences relative to those of control mice. Given these complicating factors, it is therefore important that the neurotoxicity profile of any rotenoid of therapeutic interest is established through carefully designed animal studies. A recent example of a detailed pharmacological evaluation of rotenone (1) provides a framework that can be applied to the analysis of related rotenoids.³⁶

No comprehensive study has been made of the inhibitory properties of hydroxylated rotenoids against complex I or their ability to selectively target prostate cancer cells, though a number of reports have been made on the cytotoxicity of certain hydroxylated rotenoids.^37–44 Thus, a library of 29 hydroxylated rotenoids was prepared from rotenone (1) and deguelin (2) to identify inhibitors of complex I that might be better suited to development as therapeutics than the parent compounds. The inhibitory effect of each hydroxylated rotenoid on complex I was determined and its hydrophobicity predicted. Amorphigenin (3), dalpanol (4), dihydroamorphigenin (5), and amorphigenol (6) (Figure 1) were selected for evaluation in cell-based assays using C4-2 and C4-2B bone-metastasized prostate cancer cells and control PNT2 prostate cells. The ability of each compound to reach and inhibit complex I in cells, its antiproliferative activity, and its selective toxicity towards prostate cancer cells over healthy prostate cells was then established. Each compound reaches and inhibits complex I in cells, selectively inhibits the proliferation of prostate cancer cells, and leaves healthy prostate cells unaffected. Further, the antiproliferative activities and inhibitory potencies are directly correlated. Finally, 3 and 5 are highlighted as highly active potential therapeutic candidates worthy of detailed evaluation in more sophisticated cell-based assays or animal models of prostate cancer.

Results and Discussion

Preparation of a Library of Hydroxylated Rotenoids

To prepare a library of hydroxylated derivatives of rotenone (1) and deguelin (2), redox reactions were applied to the C-12 carbonyl groups and the C-12a benzylic sites of 1 and 2, the Δ6′(7′) olefinic unit and the C-5′ and C-8′ allylic sites of 1, and the Δ4′(5′) olefinic unit of 2 (Schemes 1-4). Allylic oxidation of 1 at C-5′ with excess SeO₂ in dry 1,4-dioxane gave keto-phenol 9 via hemiacetal 9a (Scheme 1).⁴⁵ Allylic oxidation of 1 at C-8′, using N-(phenylseleno)phthalimide and H₂O in the presence of catalytic camphorsulfonic acid followed by oxidation of the resulting β-hydroxyselenide with H₂O₂ in pyridine and elimination of the tertiary selenoxide, afforded amorphigenin (3).⁴⁶ Markovnikov hydration of the Δ6′(7′) double bond, using an oxymercuration-demercuration sequence in which 1 was treated with Hg(OAc)₂ in aqueous THF followed by work-up with saturated NaCl solution and rapid demurcuration with NaBH₄ under weakly basic conditions, gave dalpanol (4).²⁷ Anti-Markovnikov hydration, using an iridium-catalyzed hydroboration-oxidation sequence in which 1 was treated with pinacolborane in the presence of catalytic [Ir(COD)Cl]₂ and bis(diphenylphosphino)ethane followed by oxidation with H₂O₂ under weakly basic conditions, provided dihydroamorphigenin (5).⁴⁷ Amorphigenol (6) was prepared from 1 as described previously.⁴⁸ Stereocontrolled reduction of the C-12-carbonyl group of 1 with NaBH₄ in MeOH afforded alcohol 10,⁴⁹ while Étard-like hydroxylation at C-12a of 1 using K₂Cr₂O₇ in aqueous AcOH provided rotenolone (7).⁵⁰ The diastereoselectivity of the last two reactions probably results from attack on the butterfly-wing conformer of 1,^49,51 with hydride delivery (as in the conversion of 1 into 10) and the pericyclic Étard-like reaction (as in the conversion of 1 into 7) taking place on the more accessible convex face of the molecule. Reduction of the C-12 carbonyl groups of 3, 4, and 6 with NaBH₄ gave diols 11 and 12, and triol 13, respectively (Scheme 1). Finally, diol 14 was obtained by concomitant hydroboration of the Δ6′(7′) double bond and reduction of the C-12 carbonyl group of 1, using BH₃.SMe₂, followed by oxidation of the intermediate borane with H₂O₂ under strongly basic conditions.

scheme 1. Synthesis of Derivatives of Rotenone (1)^a .

^aReagents and conditions: (a) NaBH₄, MeOH, 0 °C to rt, 2 h, 88-92%; (b) i. N-PSP, H₂O, CSA, CH₂Cl₂, rt, 72 h; ii. H₂O₂, pyridine 0 °C to rt, 2 h, 33%; (c) i. Hg(OAc)₂, THF, H₂O, rt, 18 h; ii. NaCl then NaHCO₃, NaBH₄, rt, 0.5 min, 48%; (d) i. [Ir(COD)Cl]₂, DPPE, HBPin, CH₂Cl₂, rt, 20 h; ii. H₂O₂, NaHCO₃, THF, H₂O, rt, 20 h, 14%; (e) OsO₄, NMO, citric acid, acetone, H₂O (see ref 48); (f) SeO₂, 4 Å MS, 1,4-dioxane, 80 °C, 3 h, 21%; (g) i. BH₃ ^.SMe₂, THF, 0 °C to rt, 1.5 h; ii. H₂O₂, NaOH, THF, H₂O, 0 °C to rt, 18 h, 80%; (h) K₂Cr₂O₇, AcOH, H₂O, 60 °C, 0.5 h then rt, 18 h, 82%. Abbreviations: MS, molecular sieves; N-PSP, N-(phenylseleno)phthalimide; CSA, (+)-β-camphorsulfonic acid; COD, 1,5-cyclooctadiene; DPPE, 1,2-bis(diphenylphosphino)ethane; HBPin, pinacolborane; NMO, N-methylmorpholine N-oxide.

scheme 4. Synthesis of Derivatives of Deguelin (2)^a .

^aReagents and conditions: (a) m-CPBA, TsOH, CHCl₃, 0 °C, 2 h, 62-69%; (b) NaBH₄, MeOH, 0 °C to rt, 2 h, 90-92%; (c) K₂Cr₂O₇, AcOH, H₂O (see ref 54); (d) OsO₄, NMO, citric acid, acetone, H₂O, rt, 28 h, 78%; (e) i. PhSeCl, CH₂Cl₂; ii. H₂O₂, THF (see ref 54). Abbreviations: m-CPBA, meta-chloroperoxybenzoic acid; TsOH, tosic acid, NMO, N-methylmorpholine N-oxide.

The library of compounds was expanded by derivatization of rotenolone (7) (Scheme 2) using the methods described above. Stereocontrolled reduction of the C-12 carbonyl group of 7 with NaBH₄ gave diol 15.⁵⁰ Regioselective allylic oxidation at C-5′ and C-8′ of 7 gave keto-phenol 16, via hemiacetal 16a, and 12a-hydroxyamorphigenin (17), respectively. Markovnikov and anti-Markovnikov hydration of the Δ6′(7′) double bond of 7 afforded 12a-hydroxydalpanol (18) and 12a-hydroxydihydroamorphigenin (19), respectively. 12a-Hydroxyamorphigenin (20) was obtained by dihydroxylation of 7 using catalytic OsO₄ and stoichiometric NMO in the presence of pyridine.⁴⁵ Reduction of 17, 18, and 20 with NaBH₄ gave triols 21 and 22, and tetraol 24, respectively. Finally, concomitant hydroboration of the Δ6′(7′) double bond and reduction of the C-12 carbonyl group of 7 followed by oxidation of the intermediate borane afforded triol 23. Reduction of ketophenols 9 (Scheme 1) and 16 (Scheme 2) was not attempted as this would have led to further rotenonic acid-like molecules, which would likely be inactive.^52,53

scheme 2. Synthesis of Derivatives of Rotenolone (7)^a .

^aReagents and conditions: (a) NaBH₄, MeOH, 0 °C to rt, 2 h, 82-90%; (b) i. N-PSP, H₂O, CSA, CH₂Cl₂, rt, 72 h; ii. H₂O₂, Al₂O₃, THF, 0 °C to rt, 6.5 h, 21%; (c) i. Hg(OAc)₂, THF, H₂O, rt, 18 h; ii. NaOH, NaBH₄, rt, 0.5 min, 51%; (d) i. [Ir(COD)Cl]₂, DPPE, HBPin, CH₂Cl₂, rt, 20 h; ii. H₂O₂, NaHCO₃, THF, H₂O, 0 °C to rt, 20 h, 12%; (e) OsO₄, NMO, pyridine, acetone, H₂O, rt, 42 h, 49%; (f) SeO₂, 4 Å MS, 1,4-dioxane, 80 °C, 3 h, 20%; (g) i. BH₃ ^.SMe₂, THF, 0 °C to rt, 1.5 h; ii. H₂O₂, NaOH, THF, H₂O, 0 °C to rt, 18 h, 87%.

Abbreviations: MS, molecular sieves; N-PSP, N-(phenylseleno)phthalimide; CSA, (+)-βcamphorsulfonic acid; COD, 1,5-cyclooctadiene; DPPE, 1,2-bis(diphenylphosphino)ethane; HBPin, pinacolborane; NMO, N-methylmorpholine N-oxide.

Next, rot-2′-enonic acid (25) was prepared from rotenone (1) as described previously (Scheme 3).⁵⁴ The isomeric rot-3′-enonic acid (26) was also prepared from 1 (Scheme 3) by ring-opening with BBr₃,^55,56 to give 4′-bromorot-2′-enonic acid (see Experimental Section), and reaction with activated zinc powder and NH₄Cl in aqueous THF.

scheme 3. Synthesis of Isomeric Rotenonic Acids^a .

^aReagents and conditions: (a) i. HBr, AcOH; ii. Zn, NH₄Cl, THF, H₂O (see ref 54); (b) BBr₃, CH₂Cl₂, −20 °C, 0.5 h, 60%; ii. Zn, NH₄Cl, THF, H₂O, rt, 0.5 h, 79%.

Finally, hydroxylated derivatives of deguelin (2), which was synthesized from 25 as described previously,⁵⁴ were prepared (Scheme 4). Epoxidation and acid-catalyzed cyclization reactions of the isomeric rotenonic acids 25 and 26, using m-CPBA and tosic acid,⁵⁷ yielded alcohols 27 and 28, via epoxides 27a and 28a, respectively. Reduction of the C-12 carbonyl group of 2 with NaBH₄ gave alcohol 29 and hydroxylation at C-12a of 2 with K₂Cr₂O₇ afforded tephrosin (8), as described previously.⁵⁴ Reduction of the C-12 carbonyl group of 8 with NaBH₄ afforded diol 30, while dihydroxylation of the Δ4′(5′) double bond of 2 using catalytic OsO₄ and stoichiometric NMO in the presence of citric acid gave diol 31.⁵⁸

In summary, by applying chemo-, regio-, and stereoselective redox reactions to rotenone (1), deguelin (2), and their derivatives, a library of 29 hydroxylated rotenoids (23 derivatives of 1 and six derivatives of 2) was assembled without the use of protecting groups.

Measurement of IC₅₀ Values and Calculation of Slog₁₀P Values for Each Derivative

Each derivative was tested for its inhibitory effect on complex I activity by using a bovine heart mitochondrial membrane assay for NADH:O₂ oxidoreduction through complexes I, III, and IV. The established bovine model for the human enzyme was used due to ease of access to material. Half-maximal inhibitory concentrations (IC₅₀ values) were calculated for all inhibitors with IC₅₀ < 10 μM (Table 1). Rotenone (1) and deguelin (2) were included as positive controls and were the most potent, with IC₅₀ values of 6.9 and 10.6 nM, respectively. Importantly, 23 out of the 29 hydroxylated derivatives (compounds 3-15, 17, 19, 23, and 25-31) were found to inhibit complex I with IC₅₀ < 10 μM. Further, where direct comparisons can be made, for compounds 1, 2, 7, and 8 for example, the IC₅₀ values agree well with those determined previously.³⁹ It should be noted that compounds 5, 6, 13, 14, 19, 20, 23, 24, 27, and 28 were tested as mixtures of two C-5′ or C-6′-diastereoisomers (epimers) due to difficulties encountered in their separation, while compound 31 was tested as a mixture of two cis-4′,5′-diastereoisomers. IC₅₀ values reported for these compounds are therefore reflective of their diastereoisomeric composition (see Experimental Section). It is well-established that the activity of rotenoids on complex I is strongly dependent on the stereochemistry of the cis-6a,12a BC ring-junction (Figure 1),^53,59,60 which is conserved within our library, and less dependent on the stereochemistry of the (remote) C-5′ substituent in the E-ring.^59,60 Individual C-5′ and, by extension, C-6′-diastereoisomers are therefore expected to have similar inhibitory activities on complex I.

Table 1. Experimentally Measured IC₅₀ Values and Computationally Predicted Slog₁₀P Values for the Rotenoids Tested.

compound	IC50 (nM)a	95 % CI	Slog10P
1	6.9	6.5-7.4	3.70
2	10.6	10.0-11.1	4.01
3	130	124-136	2.68
4	652	598-711	2.90
5b	228	220-237	2.76
6b	1800	1620-2000	1.87
7	137	127-148	3.12
8	77.5	71.6-83.8	3.42
9	2570	2460-2690	3.02
10	11.4	10.3-12.6	3.65
11	458	437-480	2.62
12	750	700-802	2.84
13b	4570	4400-4730	1.82
14b	460	425-498	2.70
15	125	119-131	3.06
16	>10000	n/a	2.44
17	3810	3520-4120	2.09
18	>10000	n/a	2.31
19b	3580	2620-5000	2.17
20b	>10000	n/a	1.29
21	>10000	n/a	2.04
22	>10000	n/a	2.26
23b	2930	2770-3090	2.12
24b	>10000	n/a	1.23
25	1090	1050-1120	4.04
26	572	540-607	4.04
27c	850	787-922	2.90
28b	195	179-213	2.90
29	14.6	13.8-15.4	3.95
30	135	128-142	3.37
31d	5310	4970-5680	2.48

^aThe error for each IC₅₀ value is reported as a 95% confidence interval (CI).

^bTested as a mixture of two C-6′-diastereoisomers (epimers).

^cTested as a mixture of two C-5′-diastereoisomers (epimers).

^dTested as a mixture of two cis-4′,5′-diastereoisomers.

Next, the octanol-water partition coefficient (Slog₁₀P) of each derivative was predicted using the Molecular Operating Environment (MOE) software package (version 2012.10, Chemical Computing Group) (Table 1). This parameter describes the hydrophobicity of a molecule and indicates the probability of it crossing the BBB by passive diffusion.^22,23 The high Slog₁₀P values for rotenone (1) and deguelin (2) of 3.70 and 4.01, respectively, were calculated for comparison, showing that they are among the most hydrophobic molecules in the library. Comparison of the IC₅₀ and Slog₁₀P data revealed a clear correlation between the hydrophobicity and inhibitory activity of the rotenoids tested, and the plot of IC₅₀ vs Slog₁₀P (Figure 2) shows that the more hydrophobic rotenoids are better complex I inhibitors. This is consistent with a binding model in which rotenoids bind in (or close to) the ubiquinone headgroup-binding site.^61,62 This binding site is accessed through a channel that leads out into the membrane,^63–66 and since rotenoids, as ubiquinone-binding site inhibitors, must partition into the lipid membrane to enter this channel, it follows that the more hydrophobic rotenoids are also the more potent inhibitors.

Figure 2.

Dependence of IC₅₀ values for complex I in bovine mitochondrial membranes on Slog₁₀P for each rotenoid tested. IC₅₀ values were determined from dose response curves measured in triplicate. The best-fit values are shown alongside error bars showing 95% confidence intervals. The dotted line at Slog₁₀P = 3 represents the cut-off value used to select rotenoids for evaluation in cell-based assays.

Selection of Hydroxylated Rotenoids for Evaluation in Cell-Based Assays of Prostate Cancer

Of the hydroxylated metabolites of rotenone (1) and deguelin (2), only rotenolone (7) and tephrosin (8), which have Slog₁₀P values of 3.12 and 3.42, respectively, are known to cross the BBB in mice and rats.²⁰ Thus, only molecules with Slog₁₀P < 3 were considered for further evaluation, as they are expected to cross the BBB less readily than 7 and 8 based on a simple hydrophobicity model (see above for a discussion of additional determinants of BBB penetration). Fourteen of the 23 inhibitors with IC₅₀ < 10 μM (3-6, 11-14, 17, 18, 23, 27, 28, and 31) have Slog₁₀P < 3. From these 14 compounds, the natural hydroxylated rotenoids amorphigenin (3), dalpanol (4), dihydroamorphigenin (5), and amorphigenol (6) were selected for evaluation in cell-based assays using PTEN-null prostate cancer cells and PTEN-positive control prostate cells. Compounds 3-6 have inhibitory potencies ranging from 130-1800 nM (Table 1) and were chosen to establish whether a correlation exists between potency and anti-proliferative effect. In addition, compounds 3-6 share a conserved core structure (the ABCD ring-system shown in Figure 1) and differ only in their E-ring substituents, so differences in their activities must rest on these structural differences alone.

Evaluation of Hydroxylated Rotenoids in Cell-Based Assays of Prostate Cancer

To establish whether amorphigenin (3), dalpanol (4), dihydroamorphigenin (5), and amorphigenol (6) reach and inhibit complex I in cells, oxygen consumption rate (OCR) assays were performed using a Seahorse XF96 instrument in which C4-2, C4-2B, and PNT2 cells were treated with each compound (1 μM). C4-2 and C4-2B cells were chosen as PTEN-null prostate cancer cells representative of advanced metastatic disease (C4-2B cells derive from C4-2 cells and therefore represent an even more invasive form of cancer). PTEN-positive PNT2 cells were selected as cells representative of normal prostate cells. All cell lines were also treated with rotenone (1) (1 μM) as a positive control. The results show that 3-6 all reach and inhibit complex I, and that the reduction in the oxygen consumption rates of all cells, independent of type, correlated with the inhibitory strength of the compound added (Figure 3a-d). Thus, cells were most affected (and most rapidly affected) by 3 (IC₅₀ 130 nM), followed by 5 (IC₅₀ 228 nM), 4 (IC₅₀ 652 nM), and then 6 (IC₅₀ 1800 nM). Interestingly, 3 was almost as effective as 1 at inhibiting the oxygen consumption of all cells, even though it is a 19-fold weaker inhibitor of complex I in membranes (Table 1), while 5 also exerted a powerful effect.

Figure 3.

Antimycin-sensitive baselined oxygen consumption rates of a) C4-2, b) C4-2B, and c) PNT2 cells measured in a Seahorse XF96 instrument. Rotenone (1), amorphigenin (3), dalpanol (4), dihydroamorphigenin (5), and amorphigenol (6) were added at 1 μM after 1 hour of basal rate measurement followed by the addition of 2 μM of rotenone and 2 μM of antimycin after 2 hours. Black; control, orange; 6, purple; 4, blue; 5, green; 3, red; 1. d) Antimycin-sentitive baselined oxygen consumption rates after 1 hour of treatment. Dark grey; C4-2, light grey; C4-2B, white; PNT2. **** P ≤ 0.0001. Error bars show S.E.M, n = 6 for treated groups, n = 16 for controls.

To determine whether the inhibition of complex I by amorphigenin (3), dalpanol (4), dihydroamorphigenin (5), and amorphigenol (6) leads to selective anti-proliferative activity in prostate cancer cells, C4-2, C4-2B, and PNT2 cells were treated with each compound (1 μM) and with rotenone (1) (1 μM) as a positive control. The results (Figure 4a-d) show that 3-6 inhibit the proliferation of the C4-2 and C4-2B prostate cancer cells, but leave the healthy PNT2 prostate cells relatively unaffected. The antiproliferative activity of each derivative against C4-2 and C4-2B was found to correlate directly with its inhibitory activity against complex I. Thus, C4-2 and C4-2B cells treated with 3 (IC₅₀ 130 nM) proliferated slowest, followed by those treated with 5 (IC₅₀ 228 nM), 4 (IC₅₀ 652 nM), and finally 6 (IC₅₀ 1800 nM). C4-2 cells were particularly affected by treatment with 3 and 5 (Figure 4a), while, encouragingly, C4-2B cells also responded to 3 and 5, but not 4 and 6 (Figure 4b). Importantly, PNT2 cells treated with 3-6 proliferated at the same rate as untreated cells (Figure 4c). C4-2 and C4-2B cells treated with 1 did not proliferate (Figure 4a-b), however, treated PNT2 cells retained their ability to proliferate, albeit at a substantially reduced rate (Figure 4c).

Figure 4.

Cell growth curves for a) C4-2, b) C4-2B, and c) PNT2 cells in the presence of 1 μM of rotenone (1), amorphigenin (3), dalpanol (4), dihydroamorphigenin (5), and amorphigenol (6). Black; control, orange; 6, purple; 4, blue; 5, green; 3, red; 1. d) Confluence (the percentage surface area of each well covered by adherent cells) after 150 hours of rotenoid treatment. Error bars show S.E.M, n = 8 for all groups. Dark grey; C4-2, light grey; C4-2B, white; PNT2. ** P ≤ 0.01, *** P ≤ 0.001,**** P ≤ 0.0001.

Finally, to confirm that the poor cell growth in the presence of rotenoids 3-6 was caused by inhibited cell proliferation rather than loss of cell viability, cells were treated for 48 hours with each compound (1 μM) before their viability was tested and cell counts were measured (Figure 5) using the acridine orange assay and a chemometec NC-3000. The data show that while rotenone (1) affected cell counts in all cells, and the cell count was lower for all compounds with C4-2 cells, no effect was seen on cell viability with any of the compounds tested. In addition, amorphigenin (3) and dihydroamorphigenin (5) had marginal effects on cell count for C4-2B cells. Therefore, under the growth conditions used (RPMI 1640 containing 11 mM glucose), inhibition of complex I leads primarily to the inhibition of cell proliferation. In contrast, a recent study reports ~70% loss of cell viability of PTEN-null cells in an MTT assay after 24 hours with deguelin (2) (0.5 μM) at the same glucose concentration.⁵ Inhibitors of the respiratory chain, including 1, have been shown to slow down the rate of the MTT to formazan reaction.⁶⁷ Interestingly, the complex I inhibitor IACS-01079 has been reported to inhibit both cancer cell proliferation and viability as assayed using Annexin V flow cytometry.¹⁹

Figure 5.

a) Cell viability, and b) cell count following 48 hours of treatment with 1 μM of rotenone (1), amorphigenin (3), dalpanol (4), dihydroamorphigenin (5), and amorphigenol (6). White, C4-2; light grey, C4-2B; dark grey, PNT2. Error bars show S.E.M, n = 4 for treated groups, n = 10 for control. * P ≤ 0.1, *** P ≤ 0.001,**** P ≤ 0.0001.

A previous study revealed that 3 and 4 are non-selectively cytotoxic towards a variety human cancer cell types, including A-549 lung, HCT-8 intestinal, RPMI-7951 skin, TE671 brain, and KB cervical cancer cells.³⁷ Half-maximal effective concentrations (ED₅₀ values) determined for all cell types were consistently lower for 3 (0.01-0.05 μg mL^-1) than for 4 (0.48-4.83 μg mL^-1),³⁷ in agreement with our data showing that in every assay 3 is more active than 4 (Figures 3-5). The cytotoxicity of several hydroxylated rotenoids towards mouse lymphoma L5178Y cells has also been reported, with IC₅₀ values ranging from 0.2-0.9 μM, as measured using the MTT assay,⁴³ suggesting that these cells are as sensitive to treatment with rotenoids as the PTEN-null prostate cancer cells used in our study. Most interestingly, a number of hydroxylated rotenoids, including rotenolone (7) and tephrosin (8) (Figure 1), have been shown to be selectively cytotoxic towards LNCaP prostate cancer cells, with ED₅₀ values ranging from 0.3-0.7 μM.⁴⁰ The mechanism of action of these compounds was not investigated, but references were made to studies on the involvement of complex I. Taken together, our results show that the inhibition of complex I is essential to the mechanism of action of 3-6, that prostate cancer cells display a greater dependence on OxPhos than healthy prostate cells, and that this dependence makes them vulnerable to inhibitors of complex I. The most significant effect of the inhibition of complex I by rotenoids 3-6 on prostate cancer cells occurs by inhibiting proliferation.

In summary, successive redox reactions were applied to rotenone (1), deguelin (2), and their derivatives to produce a library of 29 hydroxylated rotenoids. The inhibitory potency (IC₅₀) of each hydroxylated rotenoid against complex I was measured using a mitochondrial membrane assay and its hydrophobicity (Slog₁₀P) was predicted computationally. From this library, amorphigenin (3), dalpanol (4), dihydroamorphigenin (5), and amorphigenol (6) were selected for evaluation in cell-based assays using C4-2 and C4-2B prostate cancer cells and control PNT2 prostate cells. Compounds 3-6 reach and inhibit complex I in cells, selectively inhibit the proliferation of prostate cancer cells, and leave control prostate cells unaffected. Further, the clear correlation between the inhibitory potency against complex I and the anti-proliferative activity of 3-6, suggests that complex I inhibition is essential to their mechanism of action. Since the strongest antiproliferative effects were obtained with 3 and 5, these hydroxylated rotenoids, which can be synthesized as described above or extracted from natural sources,^25,28,43 are proposed as candidates for detailed evaluation in more sophisticated cell-based assays of prostate cancer and animal models. Finally, the results highlight that prostate cancer cells are more dependent on OxPhos than healthy prostate cells and that they are vulnerable to a range of rotenoid inhibitors of complex I. It is anticipated that this work will stimulate further research into the therapeutic potential of carefully selected rotenoids and related metabolism-based approaches towards the treatment of prostate cancer.

Experimental Section

General Experimental Procedures

Melting points were determined using a Büchi Melting Point B-545 melting point apparatus and are uncorrected. Optical rotations were measured using an Anton-Paar MCP 100 polarimeter. [α]_D ²⁰ values are reported in 10^-1 deg cm² g^-1 at 598 nm, concentration (c) is given in g (100 mL)^-1. Infrared spectra were recorded on a Perkin-Elmer Spectrum One spectrometer with internal referencing as neat films. Absorption maxima (ν_max) are reported in wavenumbers (cm^-1). ¹H NMR spectra were recorded on a Bruker Avance 500 Cryo Ultrashield spectrometer, operating at 500 MHz. Chemical shifts (δ) are quoted in parts per million (ppm) to the nearest 0.01 ppm and are referenced to residual solvent signals, CHCl₃ (δ 7.26) or acetone (δ 2.05). Coupling constants (J) are reported in Hertz (Hz) to the nearest 0.5 Hz. Data are reported as follows: chemical shift, integration, multiplicity (br, broad; s, singlet; d, doublet; t, triplet; q, quartet; quint, quintet; m, multiplet; app, apparent; obsc, obscured; or as combinations or these (e.g. dd)), coupling constant(s), and assignment. ¹³C NMR spectra were recorded on a Bruker Avance 500 Cryo Ultrashield spectrometer, operating at 125 MHz, with broadband proton spin decoupling. Chemical shifts (δ) are quoted in ppm to the nearest 0.1 ppm and are referenced to the solvent signals, CDCl₃ (δ 77.16) or acetone-d ₆ (δ 29.84). High resolution mass spectra were recorded using a Micromass LCT Premier spectrometer and reported mass values are within the error limits of ± 5 ppm. Analytical TLC was performed using glass plates coated with Sigma Aldrich 60 (F₂₅₄) silica gel, and compounds were visualized by ultraviolet irradiation followed by staining with ceric ammonium molybdate solution and heating. Flash chromatography was performed with Sigma-Aldrich 60 (230-400 mesh) silica gel. Rotenone (1) (Molekula Fine Chemicals, Darlington, UK, 90-95%) was crystallized from EtOH three times. Deguelin (2), amorphigenol (6), tephrosin (8), and rot-2′-enonic acid (25) were prepared from 1 as described previously.^48,54 N-(Phenylseleno)phthalimide (Sigma-Aldrich, technical grade) was crystallized from dry CH₂Cl₂/n-hexane under N₂. m-CPBA (Sigma-Aldrich, Gillingham, UK, < 77%) was crystallized from CH₂Cl₂. Zinc powder was activated by washing with 2.0 M HCl for 0.5 h, the coagulated metal was collected by filtration, washed in sequence with H₂O, EtOH, and Et₂O, dried in vacuo, and finely ground using a pestle and mortar. Powdered 4 Å molecular sieves were activated by heating in vacuo at 150 °C overnight. CH₂Cl₂, nhexane, and MeOH were distilled from CaH₂ under N₂. THF was distilled from a mixture of CaH₂ and LiAlH₄ in the presence of triphenylmethane under N₂. All other solvents and reagents were used as obtained from commercial suppliers. Moisture sensitive reactions were carried out with freshly distilled dry solvents under N₂ in glassware that was oven dried.

General Procedure 1 (GP1)

NaBH₄ (10 mol equiv) was added in small portions to a 0.02 M solution/suspension of the relevant rotenoid (1 mol equiv) in dry MeOH under N₂ at 0 °C. The mixture was warmed to rt and stirred for a further 2 h or until the reaction had gone to completion as determined by TLC. H₂O was added and the mixture was extracted with Et₂O. The organic extracts were combined, washed with brine, dried (MgSO₄), filtered, and concentrated to afford the product. Products obtained by this method were sufficiently pure such that no further purification was required.

General Procedure 2 (GP2)

BH₃ ^.SMe₂ (10 mol equiv) was added dropwise to a 0.04 M solution of the relevant rotenoid (1 mol equiv) in dry THF under N₂ at 0 °C. The mixture was stirred at 0 °C for 0.5 h, warmed to rt and stirred for a further 1 h. The mixture was cooled to 0 °C and 3.0 M aqueous NaOH solution (30 mol equiv) was added. H₂O₂ (30% aqueous solution, 30 mol equiv) was cautiously added dropwise and the mixture was stirred for 18 h while slowly warming to rt. H₂O was added and the mixture extracted with Et₂O. The organic extracts were combined, washed with brine, dried (MgSO₄), filtered, and concentrated to afford the product. Products obtained by this method were sufficiently pure such that no further purification was required.

General Procedure 3 (GP3)

Using a modified version of the procedure described for the synthesis of 9,⁴⁵ SeO₂ (4 mol equiv) was added to a 0.04 M solution/suspension of the relevant rotenoid 1 (1 mol equiv) and powdered 4 Å molecular sieves (equivalent mass to that of the rotenoid) in dry 1,4-dioxane under N₂. The mixture was heated at 80 °C for 3 h, cooled to rt, filtered through a pad of Celite, and concentrated. The residue was subjected to flash chromatography to afford the product.

General Procedure 4 (GP4)

Using the procedure described for the synthesis of 27,⁵⁷ a 0.2 M solution of m-CPBA (2 mol equiv) in dry CHCl₃ was added to a 0.04 M solution of the relevant rotenonic acid (1 mol equiv) in dry CHCl₃ under N₂ at 0 °C. Tosic acid monohydrate (0.5 mol equiv) was then added and the mixture stirred at 0 °C for 2 h. H₂O was added and the mixture extracted with CHCl₃. The organic phases were combined, washed with saturated aqueous NaHCO₃ solution and brine, dried (MgSO₄), filtered, and concentrated. The residue was subjected to flash chromatography to afford the product.

(6aS,12aS,5′R)-Amorphigenin (3) was prepared by modifying a literature procedure.⁴⁶ N-(Phenylseleno)phthalimide (153 mg, 0.508 mmol) was added to a solution of 1 (200 mg, 0.508 mmol), (+)-β-camphorsulfonic acid (12.0 mg, 0.051 mmol) and H₂O (183 μL, 10.1 mmol) in CH₂Cl₂ (10 mL) under N₂ in a flask wrapped in aluminium foil. The mixture was stirred at rt for 72 h then concentrated. The yellow solid obtained was cooled to 0 °C and dissolved in pyridine (4.0 mL) before H₂O₂ (0.5 mL, 30% aqueous solution) was added dropwise over 10 min. The mixture was stirred at 0 °C for 0.5 h, warmed to rt and stirred for 1.5 h. Saturated aqueous NaHCO₃ solution (20 mL) was added and the mixture was extracted with CHCl₃ (3 x 20 mL). The organic extracts were combined, washed with 3.0 M HCl (3 x 20 mL), H₂O (3 x 20 mL) and brine (20 mL), dried (MgSO₄), filtered, and concentrated. The residue was subjected to flash chromatography (SiO₂, 1:1 hexanes/EtOAc) to give 3 as a pale yellow solid that crystallized from CHCl₃/MeOH as colorless needles (68 mg, 33%): mp 196-198 °C (lit.⁴⁶ mp 196-197 °C);[α]_D ²⁰ −127 (c 0.1, CHCl₃) [lit.⁴⁶ [α]_D ²⁴ −124 (c 0.2, CHCl₃)]; IR ν_max 3500-3100, 1671, 1599, 1517, 1455, 1349, 1303, 1234, 1212, 1195, 1087, 817 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 3.08 (1H, dd, J 9.0, 15.5 Hz, H_a-4′), 3.39 (1H, dd, J 9.0, 15.5 Hz, H_b-4′), 3.76 (3H, s, H-2′), 3.81 (3H, s, H-3′), 3.85 (1H, d, J 4.0 Hz, H-12a), 4.18 (1H, d, J 12.0 Hz, H_a-6), 4.23 (1H, d, 13.5 Hz, H_a-8′), 4.28 (1H, d, 13.5 Hz, H_b-8′), 4.61 (1H, dd, J 3.0, 12.0 Hz, H_b-6), 4.93 (1H, dd, J 3.0, 4.0 Hz, H-6a), 5.26 (1H, s, H_a-7′), 5.28 (1H, s, H_b-7′), 5.39 (1H, t, J 9.0 Hz, H-5′), 6.45 (1H, s, H-4), 6.51 (1H, d, J 9.0 Hz, H-10), 6.76 (1H, s, H-1), 7.84 (1H, d, J 9.0 Hz, H-11); ¹³C NMR (125 MHz, CDCl₃) δ 32.0 (C-4′), 44.8 (C-12a), 56.0 (C-3′), 56.5 (C-2′), 63.1 (C-8′), 66.4 (C-6), 77.4 (C-6a), 85.7 (C-5′), 101.1 (C-4), 104.9 (C-1a), 105.1 (C-10), 110.5 (C-1), 112.9 (C-7′), 113.0 (C-8), 113.7 (C-11a), 130.2 (C-11), 144.0 (C-2), 146.8 (C-6′), 147.5 (C-3), 149.7 (C-4a), 158.1 (C-7a), 167.0 (C-9), 189.1 (C-12); HRESIMS m/z 411.1440 [M + H]⁺ (calcd for C₂₃H₂₃O₇, m/z 411.1444).

(6aS,12aS,5′R)-Dalpanol (4) was prepared by modifying a literature procedure.²⁷ [Caution: Hg(OAc)₂ and elemental Hg are highly toxic and must be handled with extreme care – all operations must be carried out in a fume-hood and all Hg-containing waste retained for proper waste disposal]. H₂O (2.0 mL) was added dropwise to a suspension of Hg(OAc)₂ (178 mg, 0.558 mmol) in THF (2.0 mL) and the bright yellow mixture was stirred at rt for 10 min. 1 (200 mg, 0.508 mmol) was added and the mixture was stirred at rt for 18 h. Brine (20 mL) was added followed by CHCl₃ (20 mL) and the two phases were thoroughly mixed. The organic layer was separated, dried (MgSO₄), filtered, and concentrated. The solid obtained was dissolved in THF (2.0 mL) and H₂O (1.5 mL). Saturated aqueous NaHCO₃ solution (0.5 mL) was added followed by NaBH₄ (19.0 mg, 0.508 mmol) and the mixture was stirred vigorously for 0.5 min as elemental Hg precipitated and coated the walls of the flask. H₂O (20 mL) was added followed by CHCl₃ (20 mL) and the two phases were thoroughly mixed. The biphasic mixture was carefully decanted and the organic layer was separated. The aqueous phase was extracted with CHCl₃ (2 x 20 mL) and the organic phases were combined, washed with brine (20 mL), dried (MgSO₄), filtered, and concentrated. The residue was subjected to flash chromatography (SiO₂, 1:1 hexanes/EtOAc) to give 4 as a white solid that crystallized from MeOH/H₂O as very fine colorless needles (128 mg, 48%): mp 208-210 °C (lit.²⁷ mp 189-191 °C, benzene); [α]_D ²⁰ −102 (c 0.1, CHCl₃) [lit.²⁷ [α]_D ²⁴ −122 (c 2.0, CHCl₃)]; IR ν_max 3500-3100, 1678, 1641, 1608, 1514, 1462, 1440, 1348, 1269, 1235, 1212, 1198, 1182, 1095, 1081, 817 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 1.22 (3H, s, H-7′), 1.35 (3H, s, H-8′), 1.76 (1H, br s, OH-6′), 3.09 (1H, dd, J 9.0, 15.0 Hz, H_a-4′), 3.14 (1H, dd, J 9.0, 15.0 Hz, H_b-4′), 3.76 (3H, s, H-2′), 3.81 (3H, s, H-3′), 3.84 (1H, d, J 4.0 Hz, H-12a), 4.18 (1H, d, J 12.0 Hz, H_a-6), 4.62 (1H, dd, J 3.0, 12.0 Hz, H_b-6), 4.67 (1H, t, J 9.0 Hz, H-5′), 4.93 (1H, dd, J 3.0, 4.0 Hz, H-6a), 6.45 (1H, s, H-4), 6.49 (1H, d, J 8.5 Hz, H-10), 6.76 (1H, s, H-1), 7.82 (1H, d, J 8.5 Hz, H-11); ¹³C NMR (125 MHz, CDCl₃) δ 24.1 (C-7′), 26.4 (C-8′), 27.5 (C-4′), 44.8 (C-12a), 56.0 (C-3′), 56.5 (C-2′), 66.4 (C-6), 71.8 (C-6a), 72.3 (C-6′), 91.6 (C-5′), 101.1 (C-4), 104.9 (C-1a), 104.9 (C-10), 110.4 (C-1), 113.6 (C-8), 113.8 (C-11a), 130.0 (C-11), 144.0 (C-2), 147.5 (C-3), 149.6 (C-4a), 158.1 (C-7a), 167.3 (C-9), 189.2 (C-12); HRESIMS m/z 413.1584 [M + H]⁺ (calcd for C₂₃H₂₅O₇, m/z 413.1595).

(6aS,12aS,5′R,6′R)- and (6aS,12aS,5′R,6′S)-6′,7′-Dihydroamorphigenin (5). 1 (200 mg, 0.508 mmol) was added to a solution of [Ir(COD)Cl]₂ (3.4 mg, 0.005 mmol), 1,2-bis(diphenylphosphino)ethane (4.0 mg, 0.010 mmol), and pinacol borane (295 μL, 2.030 mmol) in dry CH₂Cl₂ (6.0 mL) under N₂. The mixture was stirred at rt for 20 h. MeOH (1 mL) was then added and the mixture was concentrated. The resulting yellow oil was dissolved in THF (6.0 mL) and NaHCO₃ (21 mg, 0.254 mmol) was added followed by H₂O₂ (1.73 mL, 30% aqueous solution, 15.2 mmol), which was cautiously added dropwise, and the mixture stirred at rt for 20 h. H₂O (20 mL) was added and the mixture was extracted with CHCl₃ (3 x 20 mL). The organic extracts were combined, washed with brine (3 x 20 mL), dried (MgSO₄), filtered, and concentrated. The oily residue was subjected to flash chromatography (SiO₂, 1:1 hexanes/EtOAc) to give the mixture of diastereoisomers 5 as a white solid (28 mg, 14%, dr 74:26 unassigned): IR ν_max 3500-3200, 1682, 1609, 1514, 1454, 1438, 1348, 1308, 1238, 1212, 1190, 1090, 1074, 1005, 974, 829, 818 cm^-1; NMR data for the major diastereoisomer: δ _H (500 MHz, CDCl₃) 0.99 (3H, d, J 7.0 Hz, H-8′), 2.00-2.18 (1H, m obsc, H-6′), 2.95 (1H, dd, J 8.5, 16.0 Hz, H_a-4′), 3.24 (1H, dd, J 8.5, 16.0 Hz, H_b-4′), 3.66-3.74 (2H, m obsc, H_a-7′ and H_b-7′), 3.76 (3H, s, H-2′), 3.81 (3H, s, H-3"), 3.84 (1H, d, J 4.0 Hz, H-12a), 4.18 (1H, d, J 12.0 Hz, H_a-6), 4.62 (1H, dd, J 3.0, 12.0 Hz, H_b-6), 5.00 (dd, J 4.5, 8.5 Hz, H-5′), 4.93 (1H, ddd, J 1.0, 3.0, 4.0 Hz, H-6a), 6.45 (1H, s, H-4), 6.46 (1H, d, J 8.5 Hz, H-10), 6.77 (1H, s, H-1), 7.82 (1H, d, J 8.5 Hz, H-11); δ _C (125 MHz, CDCl₃) 10.9 (C-8′), 29.8 (C-4′), 40.6 (C-6′), 44.8 (C-12a), 56.0 (C-3′), 56.5 (C-2′), 65.2 (C-7′), 66.4 (C-6), 72.4 (C-6a), 86.7 (C-5′), 101.1 (C-4), 104.9 (C-1a), 105.0 (C-10), 110.5 (C-1), 113.2 (C-11a), 113.4 (C-8), 130.1 (C-11), 144.0 (C-2), 147.5 (C-4a), 149.6 (C-3), 158.1 (C-7a), 167.6 (C-9), 189.1 (C-12); NMR data for the minor diastereoisomer: δ _H (500 MHz, CDCl₃) 0.99 (3H, d, J 7.0 Hz, H-8′), 2.00-2.18 (1H, m obsc, H-6′), 2.91 (1H, dd, J 8.5, 16.0 Hz, H_a-4′), 3.27 (1H, dd, J 8.5, 16.0 Hz, H_b-4′), 3.66-3.74 (2H, m obsc, H_a-7′ and H_b-7′), 3.76 (3H, s, H-2′), 3.81 (3H, s, H-3′), 3.84 (1H, d, J 4.0 Hz, H-12a), 4.18 (1H, d, J 12.0 Hz, H_a-6), 4.62 (1H, dd, J 3.0, 12.0 Hz, H_b-6), 4.78 (dd, J 4.5, 8.5 Hz, H-5′), 4.93 (1H, ddd, J 1.0, 3.0, 4.0 Hz, H-6a), 6.45 (1H, s, H-4), 6.47 (1H, d, J 8.5 Hz, H-10), 6.77 (1H, s, H-1), 7.83 (1H, d, J 8.5 Hz, H-11); δ _C (125 MHz, CDCl₃) 12.7 (C-8′), 30.7 (C-4′), 41.1 (C-6′), 44.8 (C-12a), 56.0 (C-3′), 56.5 (C-2′), 65.2 (C-7′), 65.9 (C-6), 72.4 (C-6a), 88.9 (C-5′), 101.1 (C-4), 104.9 (C-1a), 105.1 (C-10), 110.5 (C-1), 113.2 (C-11a), 113.5 (C-8), 130.1 (C-11), 144.0 (C-2), 147.5 (C-4a), 149.6 (C-3), 158.1 (C-7a), 167.1 (C-9), 189.1 (C-12); HRESIMS m/z 435.1403 [M + Na]⁺ (calcd for C₂₃H₂₄O₇Na, m/z 435.1414).

(6aR,12aR,5′R)-Rotenolone (7) was prepared by modifying a literature procedure.⁵⁰ A solution of K₂Cr₂O₇ (2.8 g, 9.45 mmol) in H₂O (40 mL) was added dropwise over 10 min to a solution of 1 (4.0 g, 10.15 mmol) in AcOH (80 mL) at 60 °C. The mixture was stirred for 0.5 h then cooled to rt and stirred for a further 18 h. The dark green mixture was poured onto crushed ice (400 mL) and the resulting suspension was stirred for 1 h while an off-white precipitate formed. The precipitate was collected by filtration, washed thoroughly with H₂O (200 mL), dried in vacuo, and purified by flash chromatography (SiO₂, 2:1 hexanes/EtOAc) to give 7 as a white solid (3.42 g, 82%): [α]_D ²⁰ −176 (c 0.1, CHCl₃) [lit.⁵⁰ [α]_D ¹⁶ −189 (c 2.0, CHCl₃)]; IR ν_max 3600-3300, 1673, 1607, 1508, 1455, 1331, 1258, 1216, 1154, 1085, 1023, 905, 816 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 1.76 (3H, s, H-8′), 2.93 (1H, dd, J 8.0, 16.0 Hz, H_a-4′), 3.29 (1H, dd, J 8.0, 16.0 Hz, H_b-4′), 3.72 (3H, s, H-2′), 3.82 (3H, s, H-3′), 4.47 (1H, s, OH-12a), 4.49 (1H, d, J 11.5 Hz, H_a-6), 4.58 (1H, dd, J 1.0, 2.5 Hz, H-6a), 4.59 (1H, dd, J 2.5, 11.5 Hz, H_b-6), 4.93 (1H, s, H_a-7′), 5.06 (1H, s, H_b-7′), 5.23 (1H, app t, J 8.0 Hz, H-5′), 6.48 (1H, s, H-4), 6.53 (1H, d, J 8.5 Hz, H-10), 6.55 (1H, s, H-1), 7.82 (1H, d, J 8.5 Hz, H-11); ¹³C NMR (125 MHz, CDCl₃) δ 17.2 (C-8′), 31.3 (C-4′), 56.0 (C-3′), 56.5 (C-2′), 64.0 (C-6), 67.7 (C-12a), 76.2 (C-6a), 88.1 (C-5′), 101.2 (C-4), 105.5 (C-10), 108.9 (C-1a), 109.4 (C-1), 111.9 (C-11a), 112.9 (C-7′), 113.3 (C-8), 130.2 (C-11), 143.0 (C-6′), 144.1 (C-2), 148.5 (C-4a), 151.2 (C-3), 157.8 (C-7a), 168.2 (C-9), 191.2 (C-12); HRESIMS m/z 433.1241 [M + Na]⁺ (calcd for C₂₃H₂₂O₇Na, m/z 433.1258).

(6aS,12aS)-2′-Oxorot-3′-enonic Acid (9).⁴⁵ Prepared from 1 (200 mg, 0.508 mmol) by GP3 and obtained as a yellow solid (44 mg, 21%): [α]_D ²⁰ +34 (c 0.1, CHCl₃) [lit.⁴⁵ [α]_D ²⁵ +38 (c 0.1, CHCl₃)]; IR ν_max 3400-3100, 1675, 1579, 1547, 1513, 1444, 1347, 1286, 1259, 1216, 1196, 1051, 817 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 1.86 (3H, s, H-5′), 3.75 (3H, s, H-2″), 3.81 (3H, s, H-3″), 3.84 (1H, d, J 4.0 Hz, H-12a), 3.87 (1H, d, J 13.5 Hz, H_a-1′), 4.22 (1H, d, J 12.0 Hz, H_a-6), 4.23 (1H, d, J 13.5 Hz, H_b-1′), 4.64 (1H, dd, J 3.5, 12.0 Hz, H_b-6), 4.95 (1H, ddd, J 1.0, 3.5, 4.0 Hz, H-6a), 6.00 (1H, s, H_a-4′), 6.42 (1H, s, H-4), 6.59 (1H, s, H_b-4′), 6.63 (1H, d, J 8.5 Hz, H-10), 6.74 (1H, s, H-1), 7.81 (1H, d, J 8.5 Hz, H-11), 8.97 (1H, s, OH-9); ¹³C NMR (125 MHz, CDCl₃) δ 17.5 (C-5′), 32.3 (C-1′), 44.4 (C-12a), 56.1 (C-3″), 56.4 (C-2″), 66.5 (C-6), 72.6 (C-6a), 100.9 (C-4), 104.7 (C-1a), 108.5 (C-8), 110.4 (C-1), 112.6 (C-10), 112.8 (C-11a), 128.7 (C-11), 130.3 (C-4′), 143.7 (C-3′), 144.1 (C-2), 147.5 (C-4a), 149.7 (C-3), 159.1 (C-7a), 163.9 (C-9), 189.4 (C-12), 203.7 (C-2′); HRESIMS m/z 433.1239 [M + Na]⁺ (calcd for C₂₃H₂₂O₇Na, m/z 433.1258).

(6aS,12S,12aR,5′R)-12-Deoxo-12-hydroxyrotenone (10).⁴⁹ Prepared from 1 (1.0 g, 2.54 mmol) by GP1 and obtained as a white solid (921 mg, 92%): [α]_D ²⁰ −150 (c 0.1, acetone); IR ν_max 3600-3300, 1619, 1512, 1479, 1463, 1217, 1194, 1131, 1091, 1037, 799 cm^-1; ¹H NMR (500 MHz, acetone-d ₆) δ 1.76 (3H, s, H-8′), 2.88 (1H, dd, J 9.0, 15.5 Hz, H_a-4′), 3.22 (1H, dd, J 9.0, 15.5 Hz, H_b-4′), 3.44 (1H, app t, J 4.5 Hz, H-12a), 3.70 (3H, s, H-2′), 3.74 (3H, s, H-3′), 4.17 (1H, dd, J 3.0, 10.0 Hz, H_a-6), 4.48 (1H, d, J 4.0 Hz, OH-12), 4.61 (1H, app t, J 10.0 Hz, H_b-6), 4.81 (1H, ddd, J 3.0, 4.5, 10.0 Hz, H-6a), 4.88 (1H, s, H_a-7′), 5.06 (1H, s, H_b-7′), 5.12 (1H, dd, J 4.0, 4.5 Hz, H-12), 5.16 (1H, app t, J 9.0 Hz, H-5′), 6.32 (1H, d, J 8.0 Hz, H-10), 6.37 (1H, s, H-4), 7.10 (1H, d, J 8.0 Hz, H-11), 7.17 (1H, s, H-1); ¹³C NMR (125 MHz, acetone-d ₆) δ 17.4 (C-8′), 32.7 (C-4′), 38.5 (C-12a), 55.9 (C-3′), 56.8 (C-2′), 66.5 (C-6), 67.6 (C-12), 71.2 (C-6a), 86.8 (C-5′), 101.3 (C-4), 102.3 (C-10), 111.5 (C-7′), 111.6 (C-1a), 112.8 (C-8), 114.6 (C-1), 117.6 (C-11a), 129.6 (C-11), 144.4 (C-2), 145.5 (C-6′), 150.2 (C-3), 150.3 (C-4a), 150.6 (C-7a), 161.9 (C-9); HRESIMS m/z 419.1449 [M + Na]⁺ (calcd for C₂₀H₂₄O₆Na, m/z 419.1465).

(6aS,12S,12aR,5′R)-12-Deoxo-12-hydroxyamorphigenin (11). Prepared from 3 (20 mg, 0.048 mmol) by GP1 and obtained as a white solid (17.7 mg, 88%): [α]_D ²⁰ −164 (c 0.1, acetone); IR ν_max 3600-3300, 1620, 1515, 1465, 1220, 1194, 1131, 1093, 1039, 984, 787 cm^-1; ¹H NMR (500 MHz, acetone-d ₆) δ 3.00 (1H, dd, J 8.5, 15.5 Hz, H_a-4′), 3.28 (1H, dd, J 8.5, 15.5 Hz, H_b-4′), 3.44 (1H, app t, J 4.5 Hz, H-12a), 3.70 (3H, s, H-2′), 3.74 (3H, s, H-3′), 4.17 (1H, J 3.0, 9.5 Hz, H_a-6), 4.18 (1H, d, J 14.0 Hz, H_a-8′), 4.19 (1H, d, J 14.0 Hz, H_b-8′), 4.50 (1H, d, J 4.0 Hz, OH-12), 4.59 (1H, app t, J 9.5 Hz, H_b-6), 4.80 (1H, ddd, J 3.0, 4.5, 9.5 Hz, H-6a), 5.16 (1H, dd, J 4.0, 4.5 Hz, H-12), 5.17 (1H, s, H_a-7′), 5.19 (1H, s, H_b-7′), 5.30 (1H, app t, J 8.5 Hz, H-5′), 6.33 (1H, d, J 8.0 Hz, H-10), 6.37 (1H, s, H-4), 7.10 (1H, d, J 8.0 Hz, H-11), 7.17 (1H, s, H-1); ¹³C NMR (125 MHz, acetone-d ₆) δ 33.3 (C-4′), 38.5 (C-12a), 55.9 (C-3′), 56.8 (C-2′), 62.3 (C-8′), 66.5 (C-6), 67.6 (C-12), 71.2 (C-6a), 84.6 (C-5′), 101.3 (C-4), 102.3 (C-10), 109.6 (C-7′), 111.6 (C-1a), 112.9 (C-8), 114.6 (C-1), 117.7 (C-11a), 129.6 (C-11), 144.4 (C-2), 150.2 (C-6′), 150.2 (C-3), 150.3 (C-4a), 150.6 (C-7a), 161.8 (C-9); HRESIMS m/z 435.1402 [M + Na]⁺ (calcd for C₂₃H₂₄O₇Na, m/z 435.1414).

(6aS,12S,12aR,5′R)-12-Deoxo-12-hydroxydalpanol (12).²⁷ Prepared from 4 (20 mg, 0.049 mmol) by GP1 and obtained as a white solid (18.5 mg, 92%): [α]_D ²⁰ −120 (c 0.1, acetone); IR ν_max 3600-3300, 1617, 1518, 1505, 1483, 1467, 1240, 1220, 1192, 1131, 1078, 1039, 956, 789 cm^-1; ¹H NMR (500 MHz, acetone-d ₆) δ 1.20 (3H, s, H-7′), 1.24 (3H, s, H-8′), 2.98 (1H, dd, J 8.5, 13.5 Hz, H_a-4′), 3.10 (1H, dd, J 8.5, 13.5 Hz, H_b-4′), 3.43 (1H, app t, J 5.0 Hz, H-12a), 3.70 (3H, s, H-2′), 3.74 (3H, s, H-3′), 4.17 (1H, dd, J 4.0, 10.0 Hz, H_a-6), 4.44 (1H, s, OH-12), 4.59 (1H, app t, J 8.5 Hz, H-5′), 4.60 (1H, app t, J 8.5 Hz, H_b-6), 4.80 (1H, ddd, J 4.0, 5.0, 10.0 Hz, H-6a), 5.11 (1H, dd, J 4.0, 5.0 Hz, H-12), 6.26 (1H, d, J 8.0 Hz, H-10), 6.37 (1H, s, H-4), 7.05 (1H, d, J 8.0 Hz, H-11), 7.16 (1H, s, H-1); ¹³C NMR (125 MHz, acetone-d ₆) δ 25.6 (C-7′), 26.0 (C-8′), 28.5 (C-4′), 38.6 (C-12a), 55.9 (C-3′), 56.8 (C-2′), 66.5 (C-6), 67.6 (C-12), 71.1 (C-6a), 71.5 (C-6′), 90.9 (C-5′), 101.3 (C-4), 102.2 (C-10), 111.7 (C-1a), 113.7 (C-8), 114.6 (C-1), 117.3 (C-11a), 129.3 (C-11), 144.4 (C-2), 150.2 (C-4a), 150.3 (C-7a), 150.5 (C-3), 162.2 (C-9); HRESIMS m/z 437.1557 [M + Na]⁺ (calcd for C₂₃H₂₆O₇Na, m/z 437.1571).

(6aS,12S,12aR,5′R,6′R)- and (6aS,12S,12aR,5′R,6′S)-12-Deoxo-12-hydroxyamorphigenol (13). Prepared from 6 (20.0 mg, 0.047 mmol) by GP1, the mixture of diastereoisomers 13 was obtained as a white solid (17.9 mg, 89%, dr 63:37 unassigned): IR ν_max 3600-3200, 1621, 1513, 1464, 1261, 1218, 1195, 1132, 1094, 1038, 987, 787 cm^-1; NMR data for the major diastereoisomer: δ _H (500 MHz, acetone-d ₆) 1.15 (3H, s, H-8′), 2.98 (1H, dd, J 6.5, 15.5 Hz, H_a-4′), 3.17 (1H, dd, J 6.5, 15.5 Hz, H_b-4′), 3.43 (1H, app t, J 6.5 Hz, H-12a), 3.57 (1H, dd, J 5.5, 11.0 Hz, H_a-7′), 3.68 (1H, dd, J 5.5, 11.0 Hz, H_b-7′), 3.70 (3H, s, H-2′), 3.74 (3H, s, H-3′), 4.16 (1H, dd, J 3.0, 10.0 Hz, H_a-6), 4.44 (1H, d, J 4.0 Hz, OH-12), 4.60 (1H, app t, J 10.0 Hz, H_b-6), 4.81 (1H, m obsc, H-6a), 4.83 (1H, app t, J 6.0 Hz, H-5′), 5.12 (1H, dd, J 4.0, 6.5 Hz, H-12), 6.27 (1H, d, J 8.0 Hz, H-10), 6.37 (1H, s, H-4), 7.06 (1H, d, J 8.0 Hz, H-11), 7.16 (1H, s, H-1); δ _C (125 MHz, acetone-d ₆) 19.7 (C-8′), 28.0 (C-4′), 38.6 (C-12a), 55.9 (C-3′), 56.8 (C-2′), 66.5 (C-6), 67.6 (C-12), 68.1 (C-7′), 71.1 (C-6a), 73.8 (C-6′), 86.7 (C-5′), 101.4 (C-4), 102.3 (C-10), 111.7 (C-1a), 113.7 (C-8), 114.6 (C-1), 117.3 (C-11a), 129.3 (C-11), 144.4 (C-2), 150.2 (C-3), 150.3 (C-4a), 150.5 (C-7a), 162.1 (C-9); NMR data for the minor diastereoisomer: δ _H (500 MHz, acetone-d ₆) 1.16 (3H, s, H-8′), 2.98 (1H, dd, J 6.5, 15.5 Hz, H_a-4′), 3.18 (1H, dd, J 6.5, 15.5 Hz, H_b-4′), 3.43 (1H, app t, J 6.5 Hz, H-12a), 3.49 (1H, dd, J 5.5, 11.0 Hz, H_a-7′), 3.65 (1H, dd, J 5.5, 11.0 Hz, H_b-7′), 3.70 (3H, s, H-2′), 3.74 (3H, s, H-3′), 4.16 (1H, dd, J 3.0, 10.0 Hz, H_a- 6), 4.42 (1H, d, J 4.0 Hz, OH-12), 4.60 (1H, app t, J 10.0 Hz, H_b-6), 4.80 (1H, m obsc, H-6a), 4.83 (1H, app t, J 6.0 Hz, H-5′), 5.12 (1H, dd, J 4.0, 7.0 Hz, H-12), 6.25 (1H, d, J 8.0 Hz, H-10), 6.37 (1H, s, H-4), 7.04 (1H, d, J 8.0 Hz, H-11), 7.15 (1H, s, H-1); δ _C (125 MHz, acetone-d ₆) 20.8 (C-8′), 28.0 (C-4′), 38.6 (C-12a), 55.9 (C-3′), 56.8 (C-2′), 66.5 (C-6), 67.6 (C-12), 67.9 (C-7′), 71.1 (C-6a), 74.0 (C-6′), 87.8 (C-5′), 101.4 (C-4), 102.3 (C-10), 111.7 (C-1a), 113.7 (C-8), 114.6 (C-1), 117.3 (C-11a), 129.3 (C-11), 144.4 (C-2), 150.2 (C-3), 150.3 (C-4a), 150.5 (C-7a), 162.1 (C-9); HRESIMS m/z 453.1505 [M + Na]⁺ (calcd for C₂₃H₂₆O₈Na, m/z 453.1520).

(6aS,12S,12aR,5′R,6′R)- and (6aS,12S,12aR,5′R,6′S)-6′,7′-Dihydro-12-deoxo-12-hydroxyamorphigenin (14). Prepared from 1 (200 mg, 0.508 mmol) by GP2, the mixture of diastereoisomers 14 was obtained as a white solid (168 mg, 80%, dr 53:47 unassigned): IR ν_max 3600-3300, 1620, 1508, 1463, 1261, 1218, 1194, 1130, 1089, 1037, 985, 786 cm^-1; NMR data for the major diastereoisomer: δ _H (500 MHz, acetone-d ₆) 0.98 (3H, d, J 8.0 Hz, H-8′), 1.90 (1H, app quint, J 6.5, H-6′), 2.91 (1H, dd, J 9.0, 15.5 Hz, H_a-4′), 3.12 (1H, dd, J 9.0, 15.5 Hz, H_b-4′), 3.43 (1H, app t, J 6.0 Hz, H-12a), 3.52-3.57 (1H, m, H_a-7′), 3.59-3.64 (1H, m obsc, H_b-7′), 3.70 (3H, s, H-2′), 3.74 (3H, s, H-3′), 4.17 (1H, dd, J 3.0, 9.5 Hz, H_a-6), 4.42 (1H, d, J 4.5 Hz, OH-12), 4.60 (1H, app t, J 9.5 Hz, H_b-6), 4.79 (1H, m obsc, H-5′), 4.82 (1H, m obsc, H-6a), 5.11 (1H, dd, J 4.5, 5.0 Hz, H-12), 6.27 (1H, d, J 8.0 Hz, H-10), 6.37 (1H, s, H-4), 7.06 (1H, d, J 8.0 Hz, H-11), 7.15 (1H, s, H-1); δ _C (125 MHz, acetone-d ₆) 12.0 (C-8′), 31.3 (C-4′), 38.6 (C-12a), 42.2 (C-6′), 55.9 (C-3′), 56.8 (C-2′), 64.8 (C-7′), 66.5 (C-6), 67.6 (C-12), 71.1 (C-6a), 85.8 (C-5′), 101.3 (C-4), 102.2 (C-10), 111.7 (C-1a), 113.5 (C-8), 114.6 (C-1), 117.3 (C-11a), 129.5 (C-11), 144.4 (C-2), 150.2 (C-3), 150.3 (C-4a), 150.6 (C-7a), 162.2 (C-9); NMR data for the minor diastereoisomer: δ _H (500 MHz, acetone-d ₆) 0.97 (3H, d, J 8.0 Hz, H-8′), 2.04 (1H, app quint, J 6.5, H-6′), 2.86 (1H, dd, J 9.0, 15.5 Hz, H_a-4′), 3.08 (1H, dd, J 9.0, 15.5 Hz, H_b-4′), 3.43 (1H, app t, J 6.0 Hz, H-12a), 3.52-3.57 (1H, m, H_a-7′), 3.59-3.64 (1H, m obsc, H_b-7′), 3.70 (3H, s, H-2′), 3.74 (3H, s, H-3′), 4.17 (1H, dd, J 3.0, 9.5 Hz, H_a-6), 4.43 (1H, d, J 4.5 Hz, OH-12), 4.60 (1H, app t, J 9.5 Hz, H_b-6), 4.72 (1H, dd, J 8.0, 9.0 Hz, H-5′), 4.82 (1H, ddd, J 3.0, 5.0, 9.5 Hz, H-6a), 5.11 (1H, dd, J 4.5, 5.0 Hz, H-12), 6.29 (1H, d, J 8.0 Hz, H-10), 6.37 (1H, s, H-4), 7.06 (1H, d, J 8.0 Hz, H-11), 7.16 (1H, s, H-1); δ _C (125 MHz, acetone-d ₆) 12.4 (C-8′), 30.5 (C-4′), 38.6 (C-12a), 41.8 (C-6′), 55.9 (C-3′), 56.8 (C-2′), 64.8 (C-7′), 66.5 (C-6), 67.5 (C-12), 71.1 (C-6a), 86.5 (C-5′), 101.3 (C-4), 102.3 (C-10), 111.7 (C-1a), 113.3 (C-8), 114.6 (C-1), 117.4 (C-11a), 129.5 (C-11), 144.4 (C-2), 150.2 (C-3), 150.5 (C-4a), 150.6 (C-7a), 162.0 (C-9); HRESIMS m/z 437.1554 [M + Na]⁺ (calcd for C₂₃H₂₆O₇Na, m/z 437.1571).

(6aR,12R,12aS,5′R)-12-Deoxo-12-hydroxyrotenolone (15).⁵⁰ Prepared from 7 (200 mg, 0.489 mmol) by GP1 and obtained as a white solid that crystallized from MeOH/H₂O as colorless needles (173 mg, 86%): mp 126-127 °C (lit.⁵⁰ mp 125-126 °C); [α]_D ²⁰ −143 (c 0.1, CHCl₃) [lit.⁵⁰ [α]_D ²⁰ −152 (c 2.0, CHCl₃)]; IR ν_max 3600-3300, 1620, 1508, 1479, 1464, 1265, 1219, 1195, 1137, 1099, 1061, 1040, 849 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 1.78 (3H, s, H-8′), 2.14 (1H, d, J 2.5 Hz, OH-12), 2.35 (1H, s, OH-12a), 2.95 (1H, dd, J 8.5, 16.0 Hz, H_a-4′), 3.26 (1H, dd, J 8.5, 16.0 Hz, H_b-4′), 3.84 (6H, s, H-2′ and H-3′), 4.35 (1H, dd, J 4.0, 10.5 Hz, H_a-6), 4.61 (1H, dd, J 4.0, 9.5 Hz, H-6a), 4.68 (1H, dd, J 9.5, 10.5 Hz, H_b-6), 4.84 (1H, app s, H-12), 4.91 (1H, s, H_a-7′), 5.08 (1H, s, H_b-7′), 5.18 (1H, app t, J 8.5 Hz, H-5′), 6.41 (1H, s, H-4), 6.47 (1H, d, J 8.0 Hz, H-10), 7.13 (1H, d, J 8.0 Hz, H-11), 7.24 (1H, s, H-1); ¹³C NMR (125 MHz, CDCl₃) δ 17.4 (C-8′), 32.1 (C-4′), 56.0 (C-2′), 56.6 (C-3′), 65.0 (C-6), 68.6 (C-12a), 72.0 (C-12), 75.2 (C-6a), 86.8 (C-5′), 100.4 (C-4), 103.3 (C-10), 108.9 (C-1), 112.2 (C-7′), 112.3 (C-1a), 112.8 (C-8), 113.1 (C-11a), 129.7 (C-11), 143.9 (C-3), 144.1 (C-6′), 148.7 (C-7a), 149.4 (C-4a), 150.7 (C-2), 161.9 (C-9); HRESIMS m/z 435.1404 [M + Na]⁺ (calcd for C₂₃H₂₄O₇Na, m/z 435.1414).

(6aR,12aR)-2′-Oxorot-3′-enolonic Acid (16).⁴⁵ Prepared from 7 (200 mg, 0.488 mmol) by GP3 and obtained as a yellow solid (42 mg, 20%): [α]_D ²⁰ +24 (c 0.1, CHCl₃) [lit.⁴⁵ [α]_D ²⁵ +33 (c 0.1, CHCl₃)]; IR ν_max 3500-3100, 1672, 1598, 1508, 1446, 1332, 1272, 1200, 1083, 1049, 816 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 1.85 (3H, s, H-5′), 3.71 (3H, s, H-2″), 3.78 (1H, d, J 14.0 Hz, H_a-1′), 3.81 (3H, s, H-3″), 4.24 (1H, d, J 14.0 Hz, H_b-1′), 4.44 (1H, s, OH-12a), 4.53 (1H, d, J 12.0 Hz, H_a-6), 4.60 (1H, dd, J 1.0, 2.5 Hz, H-6a), 4.64 (1H, dd, J 2.5, 12.0 Hz, H_b-6), 6.00 (1H, s, H_a-4′), 6.44 (1H, s, H-4), 6.51 (1H, s, H-1), 6.55 (1H, s, H_b-4′), 6.65 (1H, d, J 8.5 Hz, H-10), 7.79 (1H, d, J 8.5 Hz, H-11) 9.13 (1H, s, OH); ¹³C NMR (125 MHz, CDCl₃) δ 17.4 (C-5′), 32.1 (C-1′), 56.1 (C-3″), 56.4 (C-2″), 64.0 (C-6), 67.4 (C-12a), 76.3 (C-6a), 101.0 (C-4), 108.5 (C-1a), 108.7 (C-8), 109.3 (C-1), 111.1 (C-11a), 113.1 (C-10), 128.7 (C-11), 130.5 (C-4′), 143.6 (C-3′), 144.2 (C-2), 148.4 (C-4a), 151.3 (C-3), 158.8 (C-7a), 164.7 (C-9), 191.4 (C-12), 203.7 (C-2′); HRESIMS m/z 449.1213 [M + Na]⁺ (calcd for C₂₃H₂₂O₈Na, m/z 449.1207).

(6aR,12aR,5′R)-12a-Hydroxyamorphigenin (17). N-(Phenylseleno)phthalimide (147 mg, 0.488 mmol) was added to solution of 7 (200 mg, 0.488 mmol), (+)-βcamphorsulfonic acid (11.0 mg, 0.049 mmol), and H₂O (176 μL, 9.76 mmol) in CH₂Cl₂ (10 mL) under N₂ in a flask wrapped in aluminium foil. The mixture was stirred at rt for 72 h and was then concentrated. The yellow solid obtained was cooled to 0 °C and dissolved in THF (4 mL) before H₂O₂ (0.5 mL, 30% aqueous solution) was added dropwise over 10 min, followed by basic Al₂O₃ (400 mg). The suspension was stirred at 0 °C for a further 0.5 h then warmed to rt and stirred for an additional 6 h. Saturated aqueous NaHCO₃ solution (20 mL) was added followed by Et₂O (20 mL) and the two phases were mixed vigorously for 10 min. The organic layer was separated, washed with H₂O (3 x 20 mL) and brine (20 mL), dried (MgSO₄), filtered, and concentrated. The residue was purified by flash chromatography (SiO₂, 1:1 hexanes/EtOAc) to give 17 as a yellow solid. The solid was dissolved in CHCl₃ and activated charcoal was added. The suspension was then filtered and concentrated to give a white solid that crystallized from MeOH/H₂O as colorless needles (44 mg, 21%): mp 89-91 °C (lit.⁶⁸ mp 92-95 °C); [α]_D ²⁰ −173 (c 0.1, CHCl₃) [lit.⁶⁸ [α]_D ²⁰ −175 (c 1.8, CHCl₃)]; IR v_max 3500-3200, 2924, 1672, 1607, 1509, 1455, 1258, 1216, 1196, 1154, 1085, 1025, 816 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 3.05 (1H, dd, J 9.0, 16.0 Hz, H_a-4′), 3.36 (1H, dd, J 9.0, 16.0 Hz, H_b-4′), 3.72 (3H, s, H-2′), 3.82 (3H, s, H-3′), 4.24 (1H, d, J 14.0 Hz, H_a-8′), 4.28 (1H, d, J 14.0 Hz, H_b-4′), 4.46 (1H, s, OH-12a), 4.48 (1H, dd, J 1.0, 11.5 Hz, H_a-6), 4.58 (1H, dd, J 1.0, 2.5 Hz, H-6a), 4.59 (1H, dd, J 2.5, 11.5 Hz, H_b-6), 5.25 (1H, s, H_a-7′), 5.28 (1H, s, H_b-7′), 5.39 (1H, app t, J 9.0 Hz, H-5′), 6.48 (1H, s, H-4), 6.53 (1H, d, J 8.5 Hz, H-10), 6.54 (1H, s, H-1), 7.83 (1H, d, J 8.5 Hz, H-11); ¹³C NMR (125 MHz, CDCl₃) δ 31.8 (C-4′), 56.0 (C-3′), 56.5 (C-2′), 63.1 (C-8′), 64.0 (C-6), 67.7 (C-12a), 76.2 (C-6a), 85.8 (C-5′), 101.2 (C-4), 105.5 (C-10), 108.8 (C-1a), 109.4 (C-1), 112.1 (C-11a), 113.0 (C-7′), 113.3 (C-8), 130.3 (C-11), 144.1 (C-2), 146.6 (C-6′), 148.5 (C-4a), 151.3 (C-3), 157.8 (C-7a), 167.7 (C-9), 191.2 (C-12); HRESIMS m/z 449.1210 [M + Na]⁺ (calcd for C₂₃H₂₂O₈Na, m/z 449.1207).

(6aR,12aR,5′R)-12a-Hydroxydalpanol (18). [Caution: Hg(OAc)₂ and elemental Hg are highly toxic and must be handled with extreme care – all operations must be carried out in a fume-hood and all Hg-containing waste retained for proper waste disposal]. H₂O (2.0 mL) was added to a suspension of Hg(OAc)₂ (171 mg, 0.537 mmol) in THF (2.0 mL) and the bright yellow solution was stirred at rt for 10 min. 7 (200 mg, 0.488 mmol) was added and the mixture was stirred at rt for 18 h. The mixture was diluted with THF (2.0 mL) and 3.0 M NaOH solution (2.0 mL) was added followed by NaBH₄ (18.5 mg, 0.488 mmol). The mixture was stirred vigorously for 0.5 min as elemental Hg rapidly precipitated and coated the walls of the flask. The reaction was worked up as described for 4 and the residue was subjected to flash chromatography (SiO₂, 1:1 hexanes/EtOAc) to give 18 as a white solid that crystallized from MeOH/H₂O as colorless needles (106 mg, 51%): mp 202-204 °C (lit.³⁷ mp 206.5-207 °C); [α]_D ²⁰ −174 (c 0.1, CHCl₃); IR ν_max 3600-3300, 2924, 1666, 1611, 1509, 1456, 1337, 1247, 1220, 1192, 1085, 1027, 951, 856, 808 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 1.21 (3H, s, H-7′), 1.34 (3H, s, H-8′), 1.76 (1H, s, OH-6′), 3.07 (1H, dd, J 9.0, 14.0 Hz, H_a-4′), 3.10 (1H, dd, J 9.0, 14.0 Hz, H_b-4′), 3.72 (3H, s, H-2′), 3.81 (3H, s, H-3′), 4.47 (1H, s, OH-12a), 4.49 (1H, dd, J 1.0, 12.0 Hz, H_a-6), 4.57 (1H, dd, J 1.0, 2.5 Hz, H-6a), 4.60 (1H, dd, J 2.5, 12.0 Hz, H_b-6), 4.68 (1H, app t, J 9.0 Hz, H-5′), 6.48 (1H, s, H-4), 6.51 (1H, d, J 8.5 Hz, H-10), 6.54 (1H, s, H-1), 7.81 (1H, d, J 8.5 Hz, H-11); ¹³C NMR (125 MHz, CDCl₃) δ 24.2 (C-7′), 26.3 (C-8′), 27.4 (C-4′), 56.0 (C-3′), 56.5 (C-2′), 64.0 (C-6), 67.7 (C-12a), 71.8 (C-6′), 76.1 (C-6a), 91.8 (C-5′), 101.2 (C-4), 105.3 (C-10), 108.8 (C-1a), 109.4 (C-1), 112.0 (C-11a), 114.0 (C-8), 130.1 (C-11), 144.1 (C-2), 148.5 (C-4a), 151.2 (C-3), 157.8 (C-7a), 167.9 (C-9), 191.2 (C-12); HRESIMS m/z 451.1367 [M + Na]⁺ (calcd for C₂₃H₂₄O₈Na, m/z 451.1363).

(6aR,12aR,5′R,6′R)- and (6aR,12aR,5′R,6′S)-6′,7′-Dihydro-12a-hydroxyamorphigenin (19). 7 (200 mg, 0.488 mmol) was added to a solution of [Ir(COD)Cl]₂ (6.5 mg, 0.010 mmol), 1,2-bis(diphenylphosphino)ethane (7.8 mg, 0.020 mmol), and pinacol borane (283 μL, 1.951 mmol) in dry CH₂Cl₂ (6.0 mL) under N₂. The mixture was stirred at rt for 20 h. MeOH (1.0 mL) was then added and the mixture was concentrated. The yellow oil obtained was dissolved in THF (6.0 mL), NaHCO₃ (20 mg, 0.254 mmol) was added followed by H₂O₂ (1.7 mL, 30% aqueous solution, 15.2 mmol), which was added dropwise over 10 min. The mixture was stirred vigorously at rt for an additional 20 h. The reaction was worked up as described for 5 and the residue was subjected to flash chromatography (SiO₂, 1:2 hexanes/EtOAc) to give the mixture of diastereoisomers 19 as a white solid (24 mg, 12%, dr 59:41 unassigned): IR ν_max 3600-3300, 2920, 1672, 1607, 1509, 1455, 1259, 1215, 1200, 1085, 1026, 817 cm^-1; NMR data for the major diastereoisomer: δ _H (500 MHz, CDCl₃) 0.98 (3H, d, J 7.0 Hz, H-8′), 2.00 (1H, dq, J 7.0, 9.0 Hz, H-6′), 2.93 (1H, dd, J 8.5, 16.0 Hz, H_a-4′), 3.22 (1H, dd, J 8.5, 16.0 Hz, H_b-4′), 3.67-3.74 (2H, m obsc, H_a-7′ and H_b-7′), 3.73 (3H, s, H-2′), 3.81 (3H, s, H-3′), 4.47 (1H, s, OH-12a), 4.49 (1H, d, J 11.5 Hz, H_a-6), 4.58 (1H, dd, J 1.5, 2.5 Hz, H-6a), 4.60 (1H, dd, J 2.5, 11.5 Hz, H_b-6), 5.00 (1H, app td, J 4.5, 9.0 Hz, H-5′), 6.48 (1H, s, H-4), 6.49 (1H, d, J 8.5 Hz, H-10), 6.54 (1H, s, H-1), 7.80 (1H, d, J 8.5 Hz, H-11); δ _C (125 MHz, CDCl₃) 10.8 (C-8′), 29.7 (C-4′), 40.6 (C-6′), 56.0 (C-3′), 56.5 (C-2′), 64.0 (C-6), 65.1 (C-7′), 67.7 (C-12a), 76.2 (C-6a), 86.7 (C-5′), 101.2 (C-4), 105.4 (C-10), 108.8 (C-1a), 109.4 (C-1), 111.7 (C-11a), 113.6 (C-8), 130.2 (C-11), 144.1 (C-2), 148.5 (C-4a), 151.3 (C-3), 157.8 (C-7a), 168.3 (C-9), 191.2 (C-12); NMR data for the minor diastereoisomer: δ _H (500 MHz, CDCl₃) 0.98 (3H, d, J 7.0 Hz, H-8′), 2.08 (1H, dq, J 7.0, 9.0 Hz, H-6′), 2.91 (1H, dd, J 8.5, 16.0 Hz, H_a-4′), 3.21 (1H, dd, J 8.5, 16.0 Hz, H_b-4′), 3.67-3.74 (2H, m obsc, H_a-7′ and H_b-7′), 3.73 (3H, s, H-2′), 3.81 (3H, s, H-3′), 4.47 (1H, s, OH-12a), 4.49 (1H, d, J 11.5 Hz, H_a-6), 4.58 (1H, dd, J 1.5, 2.5 Hz, H-6a), 4.60 (1H, dd, J 2.5, 11.5 Hz, H_b-6), 4.76 (1H, dd, J 4.5, 9.0 Hz, H-5′), 6.48 (1H, s, H-4), 6.49 (1H, d, J 8.5 Hz, H-10), 6.54 (1H, s, H-1), 7.80 (1H, d, J 8.5 Hz, H-11); δ _C (125 MHz, CDCl₃) 12.6 (C-8′), 30.5 (C-4′), 41.0 (C-6′), 56.0 (C-3′), 56.5 (C-2′), 64.0 (C-6), 65.7 (C-7′), 67.7 (C-12a), 76.2 (C-6a), 88.9 (C-5′), 101.2 (C-4), 105.5 (C-10), 108.8 (C-1a), 109.4 (C-1), 111.8 (C-11a), 113.4 (C-8), 130.2 (C-11), 144.1 (C-2), 148.5 (C-4a), 151.3 (C-3), 157.8 (C-7a), 167.8 (C-9), 191.2 (C-12); HRESIMS m/z 451.1350 [M + Na]⁺ (calcd for C₂₃H₂₄O₈Na, m/z 451.1363).

(6aR,12aR,5′R,6′R)- and (6aR,12aR,5′R,6′S)-12a-Hydroxyamorphigenol (20) were prepared by modifying a literature procedure.⁴⁵ OsO₄ (99 μL of a 2.5 wt% solution in t-BuOH, 0.010 mmol) was added to a solution of 7 (200 mg, 0.488 mmol), N-methylmorpholine N-oxide (71 mg, 0.976 mmol), and pyridine (78 μL, 0.976 mmol) in acetone (8.0 mL) and H₂O (0.8 mL). The mixture was stirred at rt for 42 h. EtOAc (20 mL) was added followed by saturated aqueous Na₂SO₃ solution (20 mL) and the two phases were mixed vigorously for 10 min. The organic layer was separated, washed with 3.0 M HCl (3 x 20 mL), H₂O (20 mL) and brine (20 mL), dried (MgSO₄), filtered, and concentrated. The residue was subjected to flash chromatography (SiO₂, 1:1 hexanes/EtOAc) to give the mixture of diastereoisomers 20 as a white solid (106 mg, 49%, dr 62:38 unassigned): IR ν_max 3600-3300, 1672, 1607, 1509, 1455, 1338, 1248, 1216, 1199, 1155, 1085, 1048, 1025, 817, 746 cm^-1; NMR data for the major diastereoisomer: δ _H (500 MHz, CDCl₃) 1.21 (3H, s, H-8′), 3.17 (1H, dd, J 11.0, 17.5 Hz, H_a-4′), 3.19 (1H, dd, J 11.0, 17.5 Hz, H_b-4′), 3.53 (1H, d, J 11.0 Hz, H_a-7′), 3.71 (3H, s, H-2′), 3.74 (1H, d, J 11.0 Hz, H_b-7′), 3.82 (3H, s, H-3′), 4.46 (1H, s, OH-12a), 4.49 (1H, d, J 12.0 Hz, H_a-6), 4.59 (1H, dd, 1.0, 2.5 Hz, H-6a), 4.60 (1H, dd, J 2.5, 12.0 Hz, H_b-6), 4.86 (1H, app t, J 11.0 Hz, H-5′), 6.49 (1H, s, H-4), 6.50 (1H, d, J 8.5 Hz, H-10), 6.53 (1H, s, H-1), 7.80 (1H, d, J 8.5 Hz, H-11); δ _C (125 MHz, CDCl₃) 19.4 (C-8′), 27.0 (C-4′), 56.0 (C-3′), 56.5 (C-2′), 63.9 (C-6), 67.0 (C-7′), 67.8 (C-12a), 73.5 (C-6′), 76.2 (C-6a), 87.5 (C-5′), 101.2 (C-4), 105.4 (C-10), 108.8 (C-1a), 109.4 (C-1), 112.1 (C-11a), 113.9 (C-8), 130.1 (C-11), 144.1 (C-2), 148.6 (C-4a), 151.3 (C-3), 157.8 (C-7a), 167.7 (C-9), 191.3 (C-12); NMR data for the minor diastereoisomer: δ _H (500 MHz, CDCl₃) 1.16 (3H, s, H-8′), 3.16 (1H, dd, J 11.0, 17.5 Hz, H_a-4′), 3.20 (1H, dd, J 11.0, 17.5 Hz, H_b-4′), 3.57 (1H, d, J 11.0 Hz, H_a-7′), 3.71 (3H, s, H-2′), 3.79 (1H, d, J 11.0 Hz, H_b-7′), 3.82 (3H, s, H-3′), 4.45 (1H, s, OH-12a), 4.49 (1H, d, J 12.0 Hz, H_a-6), 4.59 (1H, dd, 1.0, 2.5 Hz, H-6a), 4.60 (1H, dd, J 2.5, 12.0 Hz, H_b-6), 4.86 (1H, app t, J 11.0 Hz, H-5′), 6.48 (1H, s, H-4), 6.51 (1H, d, J 8.5 Hz, H-10), 6.53 (1H, s, H-1), 7.81 (1H, d, J 8.5 Hz, H-11); δ _C (125 MHz, CDCl₃) 19.8 (C-8′), 27.3 (C-4′), 56.0 (C-3′), 56.5 (C-2′), 63.9 (C-6), 67.7 (C-12a), 68.6 (C-7′), 73.1 (C-6′), 76.2 (C-6a), 90.2 (C-5′), 101.2 (C-4), 105.3 (C-10), 108.8 (C-1a), 109.4 (C-1), 112.2 (C-11a), 113.7 (C-8), 130.1 (C-11), 144.1 (C-2), 148.6 (C-4a), 151.3 (C-3), 157.8 (C-7a), 167.4 (C-9), 191.3 (C-12); HRESIMS m/z 467.1316 [M + Na]⁺ (calcd for C₂₃H₂₄O₉Na, m/z 467.1313).

(6aR,12R,12aS,5′R)-12-Deoxo-12,12a-dihydroxyamorphigenin (21).⁶⁹ Prepared from 17 (18.0 mg, 0.042 mmol) by GP1 and obtained as a white solid (14.8 mg, 82%): [α]_D ²⁰ −86 (c 0.1, CHCl₃); IR ν_max 3600-3300, 1620, 1509, 1466, 1265, 1221, 1194, 1137, 1099, 1036, 734 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 2.25 (1H, d, J 2.5 Hz, OH-12), 2.51 (1H, s, OH-12a), 3.04 (1H, dd, J 9.0, 15.5 Hz, H_a-4′), 3.32 (1H, dd, J 9.0, 15.5 Hz, H_b-4′), 3.83 (6H, s, H-2′ and H-3′), 4.22 (1H, d, J 14.0 Hz, H_a-8′), 4.27 (1H, d, J 14.0 Hz, H_b-8′), 4.32 (1H, dd, J 4.0, 10.5 Hz, H_a-6), 4.59 (1H, dd, J 4.0, 8.0 Hz, H-6a), 4.66 (1H, dd, J 8.0, 10.5 Hz, H_b-6), 4.84 (1H, app s, H-12), 5.24 (1H, s, H_a-7′), 5.25 (1H, s, H_b-7′), 5.33 (1H, app t, J 9.0 Hz, H-5′), 6.40 (1H, s, H-4), 6.46 (1H, d, J 8.5 Hz, H-10), 7.14 (1H, d, J 8.5 Hz, H-11), 7.24 (1H, s, H-1); ¹³C NMR (125 MHz, CDCl₃) δ 32.6 (C-4′), 56.0 (C-2′), 56.6 (C-3′), 63.2 (C-8′), 64.9 (C-6), 68.5 (C-12a), 72.0 (C-12), 75.2 (C-6a), 84.8 (C-5′), 100.4 (C-4), 103.3 (C-10), 109.0 (C-1), 112.3 (C-1a), 112.6 (C-8), 112.7 (C-7′), 113.6 (C-11a), 129.7 (C-11), 144.1 (C-3), 147.4 (C-6′), 148.7 (C-7a), 149.4 (C-4a), 150.7 (C-2), 161.3 (C-9); HRESIMS m/z 451.1373 [M + Na]⁺ (calcd for C₂₃H₂₄O₈Na, m/z 451.1369.

(6aR,12R,12aS,5′R)-12-Deoxo-12,12a-dihydroxydalpanol (22). Prepared from 18 (22.0 mg, 0.051 mmol) by GP1 and obtained as a colorless oil that solidified upon standing at rt to a white solid (19.9 mg, 90%): [α]_D ²⁰ −134 (c 0.1, CHCl₃); IR ν_max 3500-3200, 1620, 1508, 1481, 1464, 1266, 1220, 1194, 1138, 1098, 1063, 1038, 968 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 1.20 (3H, s, H-7′), 1.33 (3H, s, H_a-8′), 1.94 (1H, s, OH-6′), 2.37 (1H, d, J 2.0 Hz, OH-12), 2.67 (1H, s, OH-12a), 3.03 (1H, dd, J 9.5, 17.0 Hz, H_a-4′), 3.06 (1H, dd, J 9.5, 17.0 Hz, H_b-4′), 3.81 (3H, s, H-2′), 3.82 (3H, s, H-3′), 4.34 (1H, dd, J 4.0, 10.5 Hz, H_a-6), 4.51 (1H, dd, J 4.0, 9.0 Hz, H-6a), 4.52 (1H, m obsc, H-5′), 4.65 (1H, dd, J 9.0, 10.5 Hz, H_b-6), 4.82 (1H, app s, H-12), 6.38 (1H, s, H-4), 6.42 (1H, d, J 8.0 Hz, H-10), 7.11 (1H, d, J 8.0 Hz, H-11), 7.24 (1H, s, H-1); ¹³C NMR (125 MHz, CDCl₃) δ 24.0 (C-7′), 26.4 (C-8′), 28.1 (C-4′), 56.0 (C-2′), 56.6 (C-3′), 64.9 (C-6), 68.4 (C-12a), 72.0 (C-6′), 72.1 (C-12), 75.2 (C-6a), 90.4 (C-5′), 100.4 (C-4), 103.0 (C-10), 109.2 (C-1), 112.4 (C-1a), 113.3 (C-8), 113.5 (C-11a), 129.4 (C-11), 144.0 (C-3), 148.7 (C-7a), 149.4 (C-4a), 150.6 (C-2), 161.5 (C-9); HRESIMS m/z 435.1502 [M + Na]⁺ (calcd for C₂₃H₂₆O₈Na, m/z 453.1520).

(6aR,12R,12aS,5′R,6′R)- and (6aR,12R,12aS,5′R,6′S)-6′,7′-Dihydro-12-deoxo-12,12a-dihydroxyamorphigenin (23). Prepared from 7 (200 mg, 0.488 mmol) by GP2, the mixture of diastereoisomers 23 was obtained as a white solid (182 mg, 87%, dr 57:43 unassigned): IR ν_max 3600-3300, 1619, 1512, 1464, 1281, 1218, 1193, 1130, 1090, 985, 870, 786 cm^-1; NMR data for the major diastereoisomer: δ _H (500 MHz, CDCl₃) 0.99 (3H, d, J 7.0 Hz, H-8′), 2.02 (1H, m, H-6′), 2.32 (1H, s, OH-12), 2.63 (1H, s, OH-12a), 2.92 (1H, dd, J 9.0, 15.0 Hz, H_a-4′), 3.16 (1H, dd, J 9.0, 15.0 Hz, H_b-4′), 3.64 (1H, dd, J 4.5, 11.0 Hz, H_a-7′), 3.74 (1H, dd, J 4.5, 11.0 Hz, H_b-7′), 3.82 (6H, s, H-2′ and H-3′), 4.34 (1H, dd, J 4.0, 10.5 Hz, H_a-6), 4.60 (1H, dd, J 4.0, 9.0 Hz, H-6a), 4.68 (1H, dd, J 9.0, 10.5 Hz, H_b-6), 4.83 (1H, app s, H-12), 4.88 (1H, app td, J 4.5, 9.0 Hz, H-5′), 6.39 (1H, s, H-4), 6.41 (1H, d, J 8.0 Hz, H-10), 7.11 (1H, d, J 8.0 Hz, H-11), 7.23 (1H, s, H-1); δ _C (125 MHz, CDCl₃) 11.2 (C-8′), 30.1 (C-4′), 40.3 (C-6′), 56.0 (C-2′), 56.6 (C-3′), 64.9 (C-6), 65.4 (C-7′), 68.5 (C-12a), 72.0 (C-12), 75.2 (C-6a), 85.7 (C-5′), 100.3 (C-4), 103.1 (C-10), 109.0 (C-1), 112.4 (C-1a), 113.0 (C-8), 113.2 (C-11a), 129.6 (C-11), 144.0 (C-3), 148.7 (C-7a), 149.4 (C-4a), 150.6 (C-2), 161.7 (C-9); NMR data for the minor diastereoisomer: δ _H (500 MHz, CDCl₃) 0.95 (3H, d, J 7.0 Hz, H-8′), 2.02 (1H, m, H-6′), 2.32 (1H, s, OH-12), 2.63 (1H, s, OH-12a), 2.90 (1H, dd, J 9.0, 15.0 Hz, H_a-4′), 3.18 (1H, dd, J 9.0, 15.0 Hz, H_b-4′), 3.66 (1H, dd, J 4.5, 11.0 Hz, H_a-7′), 3.72 (1H, dd, J 4.5, 11.0 Hz, H_b-7′), 3.82 (6H, s, H-2′ and H-3′), 4.34 (1H, dd, J 4.0, 10.5 Hz, H_a-6), 4.60 (1H, dd, J 4.0, 9.0 Hz, H-6a), 4.63 (1H, dd, J 4.5, 9.0 Hz, H-5′), 4.69 (1H, dd, J 9.0, 10.5 Hz, H_b-6), 4.83 (1H, app s, H-12), 6.39 (1H, s, H-4), 6.41 (1H, d, J 8.0 Hz, H-10), 7.12 (1H, d, J 8.0 Hz, H-11), 7.23 (1H, s, H-1); δ _C (125 MHz, CDCl₃) 12.9 (C-8′), 31.6 (C-4′), 41.1 (C-6′), 56.0 (C-2′), 56.6 (C-3′), 64.9 (C-6), 65.4 (C-7′), 68.5 (C-12a), 72.0 (C-12), 75.2 (C-6a), 88.4 (C-5′), 100.3 (C-4), 103.2 (C-10), 109.0 (C-1), 112.4 (C-1a), 112.7 (C-8), 113.5 (C-11a), 129.6 (C-11), 144.0 (C-3), 148.7 (C-7a), 149.4 (C-4a), 150.6 (C-2), 161.2 (C-9); HRESIMS m/z 453.1524 [M + Na]⁺ (calcd for C₂₃H₂₆O₈Na, m/z 453.1520).

(6aR,12R,12aS,5′R,6′R)- and (6aR,12R,12aS,5′R,6′S)-12-Deoxo-12,12a-dihydroxyamorphigenol (24). Prepared from 20 (20.0 mg, 0.045 mmol) by GP1, the mixture of diastereoisomers 24 was obtained as a white solid (18.1 mg, 90%, dr 62:38 unassigned): IR ν_max 3600-3300, 1621, 1515, 1479, 1219, 1195, 1132, 1094, 1039, 801 cm-¹; NMR data for the major diastereoisomer: δ _H (500 MHz, CDCl₃) 1.22 (3H, s, H-8′), 2.39 (1H, s, OH-12), 2.48 (1H, s, OH-12a), 3.13 (1H, dd, J 9.0, 15.5 Hz, H_a-4′), 3.15 (1H, dd, J 9.0, 15.5 Hz, H_b-4′), 3.52 (1H, d, J 13.0 Hz, H_a-7′), 3.75 (1H, d, J 13.0 Hz, H_b-7′), 3.83 (6H, s, H-2′ and H-3′), 4.36 (1H, dd, J 3.5, 10.5 Hz, H_a-6), 4.62 (1H, dd, J 3.5, 9.0 Hz, H-6a), 4.68 (1H, dd, J 9.0, 10.5 Hz, H_b-6), 4.77 (app t, J 9.0 Hz, H-5′), 4.84 (1H, app s, H-12), 6.41 (1H, s, H-4), 6.43 (1H, d, J 8.0 Hz, H-10), 7.12 (1H, d, J 8.0 Hz, H-11), 7.26 (1H, s, H-1); δ _C (125 MHz, CDCl₃) 19.8 (C-8′), 27.8 (C-4′), 56.0 (C-2′), 56.6 (C-3′), 64.9 (C-6), 67.0 (C-7′), 68.5 (C-12a), 72.0 (C-12), 73.7 (C-6′), 75.2 (C-6a), 87.0 (C-5′), 100.4 (C-4), 103.1 (C-10), 109.0 (C-1), 112.3 (C-1a), 113.1 (C-8), 113.6 (C-11a), 129.6 (C-11), 144.1 (C-3), 148.8 (C-7a), 149.4 (C-4a), 150.7 (C-2), 161.3 (C-9); NMR data for the minor diastereoisomer: δ _H (500 MHz, CDCl₃) 1.15 (3H, s, H-8′), 2.39 (1H, s, OH-12), 2.58 (1H, s, OH-12a), 3.12 (1H, dd, J 9.0, 15.5 Hz, H_a-4′), 3.22 (1H, dd, J 9.0, 15.5 Hz, H_b-4′), 3.54 (1H, d, J 13.0 Hz, H_a-7′), 3.77 (1H, d, J 13.0 Hz, H_b-7′), 3.83 (6H, s, H-2′ and H-3′), 4.36 (1H, dd, J 3.5, 10.5 Hz, H_a-6), 4.62 (1H, dd, J 3.5, 9.0 Hz, H-6a), 4.68 (1H, dd, J 9.0, 10.5 Hz, H_b-6), 4.78 (app t, J 9.0 Hz, H-5′), 4.84 (1H, app s, H-12), 6.41 (1H, s, H-4), 6.43 (1H, d, J 8.0 Hz, H-10), 7.12 (1H, d, J 8.0 Hz, H-11), 7.25 (1H, s, H-1); δ _C (125 MHz, CDCl₃) 19.8 (C-8′), 28.0 (C-4′), 56.0 (C-2′), 56.6 (C-3′), 64.9 (C-6), 68.4 (C-12a), 68.7 (C-7′), 72.0 (C-12), 73.1 (C-6′), 75.2 (C-6a), 89.1 (C-5′), 100.4 (C-4), 103.2 (C-10), 109.0 (C-1), 112.2 (C-1a), 112.9 (C-8), 113.9 (C-11a), 129.5 (C-11), 144.1 (C-3), 148.8 (C-7a), 149.4 (C-4a), 150.7 (C-2), 161.3 (C-9); HRESIMS m/z 469.1451 [M + Na]⁺ (calcd for C₂₃H₂₆O₉Na, m/z 469.1469).

(6aS,12aS)-4′-Bromorot-2′-enonic Acid was prepared by adapting several literature procedures.^55,56 The following procedure gave consistently reproducible results: BBr₃ (2.54 mL of a 1.0 M solution in CH₂Cl₂, 2.54 mmol) was added dropwise over 10 min to a solution of 1 (1.00 g, 2.54 mmol) in dry CH₂Cl₂ (6.0 mL) under N₂ at −20 °C. The mixture was stirred at −20 °C for a further 20 min before the reaction was quenched with H₂O (20 mL), quickly warmed to rt, and extracted with EtOAc (20 mL). The organic layer was separated, washed with H₂O (3 x 20 mL) and brine (20 mL), dried (MgSO₄), filtered, and concentrated. The yellow oily residue was dissolved in MeOH (14 mL) and the solution set aside at 4 °C overnight to give 4′-bromorot-2′-enonic acid as a white powdery solid that was collected by filtration and dried in vacuo. The solid crystallized from CH₂Cl₂/MeOH at 0 °C as large colorless needles (0.718 g, 60%): mp 152-154 °C (lit.⁷⁰ mp 152-154 °C); [α]_D ²⁰ +22 (c 0.1, CHCl₃) [lit.⁷⁰ [α]_D ²⁰ +27 (c 0.7, CHCl₃)]; IR ν_max 3200-3000, 1662, 1595, 1513, 1453, 1442, 1350, 1276, 1214, 1199, 1098, 882, 814 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 1.93 (3H, s, H-5′), 3.37 (1H, dd, J 7.5, 15.0 Hz, H_a-1′), 3.42 (1H, dd J 7.5, 15.0 Hz, H_b-1′), 3.75 (3H, s, H-2″), 3.80 (3H, s, H-3″), 3.83 (1H, d, J 4.0 Hz, H-12a), 3.95 (2H, s, H-4′), 4.18 (1H, d, J 12.0 Hz, H_a-6), 4.62 (1H, dd, J 3.5, 12.0 Hz, H_b-6), 4.91 (1H, dd, J 3.5, 4.0 Hz, H-6a), 5.63 (1H, t, J 7.5 Hz, H-2′), 5.93 (1H, s, OH-9), 6.45 (1H, s, H-4), 6.52 (1H, d, J 8.5 Hz, H-10), 6.76 (1H, s, H-1), 7.74 (1H, d, J 8.5 Hz, H-11); ¹³C NMR (125 MHz, CDCl₃) δ 14.9 (C-5′), 22.5 (C-1′), 41.5 (C-4′), 44.4 (C-12a), 56.0 (C-3″), 56.5 (C-2″), 66.4 (C-6), 72.4 (C-6a), 101.1 (C-4), 104.7 (C-1a), 110.6 (C-1), 110.8 (C-10), 113.3 (C-8), 113.9 (C-11a), 127.5 (C-11), 128.0 (C-2′), 133.4 (C-3′), 143.9 (C-2), 147.7 (C-3), 149.6 (C-4a), 160.4 (C-7a), 161.0 (C-9), 189.9 (C-12); HRESIMS m/z 475.0739 [M + H]⁺ (calcd for C₂₃H₂₄O₆ ⁷⁹Br, m/z 475.0756) and m/z 477.039 [M + H]⁺ (calcd for C₂₃H₂₄O₆ ⁸¹Br, m/z 477.0736).

(6aS,12aS)-Rot-3′-enonic Acid (26). Zinc powder (219 mg, 3.368 mmol, activated as described above) was added to a solution of 4′-bromorot-2′-enonic acid (400 mg, 0.842 mmol) and NH₄Cl (180, mg, 3.368 mmol) in THF (12 mL) and H₂O (2.0 mL) and the suspension was stirred vigorously at rt for 0.5 h. The mixture was filtered through a pad of Celite and the inorganic residues were washed with EtOAc (80 mL). The combined filtrates were washed with H₂O (3 x 80 mL) and brine (80 mL), dried (MgSO₄), filtered, and concentrated to afford 26 as a pale yellow solid that crystallized from MeOH/H₂O as small colorless prisms (264 mg, 79%): mp 177-178 °C (lit.⁷¹ 176-178 °C); [α]_D ²⁰ +69 (c 0.1, CHCl₃); IR ν_max 3600-3300, 1657, 1604, 1594, 1515, 1447, 1439, 1350, 1296, 1273, 1216, 1192, 1077, 1006, 814 cm^-1; ¹H NMR (500 MHz, CDCl₃) δ 1.78 (3H, s, H-5′), 2.19 (2H, ddd, J 5.0, 7.0, 11.0 Hz, H-2′), 2.73 (1H, ddd, J 5.0, 7.0, 10.5 Hz, H_a-1′), 2.81 (1H, ddd, J 5.0, 7.0, 10.5 Hz, H_b-1′), 3.75 (3H, s, H-2″), 3.80 (3H, s, H-3″), 3.83 (1H, d, J 3.0 Hz, H-12a), 4.17 (1H, d, J 10.0 Hz, H_a-6), 4.60 (1H, d, J 1.0 Hz, H_a-4′), 4.62 (1H, dd, J 3.0, 10.0 Hz, H_b-6), 4.66 (1H, d, J 1.0 Hz, H_b-4′), 4.88 (1H, ddd, J 1.0, 2.5, 3.0 Hz, H-6a), 5.88 (1H, s, OH-9), 6.45 (1H, s, H-4), 6.47 (1H, d, J 7.0 Hz, H-10), 6.77 (1H, s, H-1), 7.72 (1H, d, J 7.0 Hz, H-11); ¹³C NMR (125 MHz, CDCl₃) δ 21.6 (C-1′), 22.6 (C-5′), 36.9 (C-2′), 44.4 (C-12a), 56.0 (C-3"), 56.5 (C-2"), 66.5 (C-6), 72.3 (C-6a), 101.0 (C-4), 104.9 (C-1a), 110.5 (C-4′), 110.5 (C-10), 110.7 (C-1), 113.1 (C-8), 116.1 (C-11a), 127.0 (C-11), 143.9 (C-2), 145.9 (C-3′), 147.7 (C-3), 149.6 (C-4a), 160.7 (C-7a), 160.9 (C-9), 190.2 (C-12); HRESIMS m/z 397.1636 [M + H]⁺ (calcd for C₂₀H₁₆O₆, m/z 397.1651).

(6aS,12aS,5′R)- and (6aS,12aS,5′S)-4′,5′-Dihydro-5′-hydroxydeguelin (27).⁵⁷ Prepared from 25 (180 mg, 0.455 mmol) by GP4, the mixture of diastereoisomers 27 was obtained as a white solid (116 mg, 62%, dr 57:43 unassigned): IR ν_max 3600-3300, 1665, 1601, 1582, 1512, 1440, 1342, 1261, 1212, 1195, 1136, 1094, 1043, 1008, 817 cm^-1; NMR data for the major diasteroisomer: δ _H (500 MHz, CDCl₃) 1.32 (3H, s, H-7′), 1.34 (3H, s, H-8′), 1.65 (1H, s, OH-5′), 2.69 (1H, dd, J 5.5, 16.0 Hz, H_a-4′), 2.95 (1H, dd, J 5.5, 16.0 Hz, H_b-4′) 3.77 (3H, s, H-2′), 3.81 (3H, s, H-3′), 3.83-3.90 (2H, m obsc, H-12a and H-5′), 4.19 (1H, d, J 11.0 Hz, H_a-6), 4.63 (1H, dd, J 3.5, 11.0 Hz, H_b-6), 4.92 (1H, ddd, J 1.0, 3.5, 4.0 Hz, H-6a), 6.45 (1H, s, H-4) 6.50 (1H, d, J 9.0 Hz, H-10) 6.79 (1H, s, H-1), 7.75 (1H, d, J 9.0 Hz, H-11); δ _C (125 MHz, CDCl₃) 21.8 (C-7′), 25.1 (C-8′), 26.0 (C-4′), 44.4 (C-12a), 56.0 (C-3′), 56.5 (C-2′), 66.4 (C-6), 69.1 (C-5′), 72.6 (C-6a), 78.3 (C-6′), 101.1 (C-4), 104.9 (C-1a), 107.2 (C-8), 110.5 (C-1), 112.2 (C-10), 112.4 (C-11a), 127.1 (C-11), 144.0 (C-2), 147.6 (C-4a), 149.6 (C-3), 160.0 (C-9), 160.5 (C-7a), 189.5 (C-12); NMR data for the minor diasteroisomer: δ _H (500 MHz, CDCl₃) 1.27 (3H, s, H-7′) 1.39 (3H, s, H-8′) 1.72 (1H, s, OH-5′), 2.77 (1H, dd, J 5.5, 16.0 Hz, H_a-4′) 2.90 (1H, dd, J 5.5, 16.0 Hz, H_b-4′), 3.77 (3H, s, H-2′), 3.81 (3H, s, H-3′), 3.83-3.90 (2H, m obsc, H-12a and H-5′) 4.19 (1H, d, J 11.0 Hz, H_a-6), 4.64 (1H, dd, J 3.5, 11.0 Hz, H_b-6), 4.92 (1H, ddd, J 1.0, 3.5, 4.0 Hz, H-6a), 6.45 (1H, s, H-4), 6.51 (1H, d, J 9.0 Hz, H-10) 6.80 (1H, s, H-1), 7.76 (1H, d, J 9.0 Hz, H-11); δ _C (125 MHz, CDCl₃) 22.7 (C-7′), 24.8 (C-8′), 26.0 (C-4′), 44.4 (C-12a), 56.0 (C-3′), 56.5 (C-2′), 66.5 (C-6), 68.7 (C-5′), 72.6 (C-6a), 78.2 (C-6′), 101.1 (C-4), 105.0 (C-1a), 106.9 (C-8), 110.5 (C-1), 112.3 (C-10), 112.5 (C-11a), 127.1 (C-11), 144.0 (C-2), 147.6 (C-4a), 149.6 (C-3), 160.0 (C-9), 160.7 (C-7a), 189.5 (C-12); HRESIMS m/z 435.1400 [M + Na]⁺ (calcd for C₂₃H₂₄O₇Na, m/z 435.1414).

(6aS,12aS,6′R)- and (6aS,12aS,6′S)-4′,5′-Dihydro-7′-hydroxydeguelin (28). Prepared from 26 (180 mg, 0.455 mmol) by GP4, the mixture of diastereoisomers 28 was obtained as a white solid (129 mg, 69%, dr 53:47 unassigned): IR ν_max 3600-3300, 1664, 1599, 1580, 1511, 1439, 1335, 1249, 1212, 1196, 1093, 1056, 1003, 817 cm^-1; NMR data for the major diasteroisomer: δ _H (500 MHz, CDCl₃) 1.22 (3H, s, H-8′), 1.70-1.75 (1H, m obsc, H_a-5′), 1.86 (1H, s, OH-7′), 1.93-1.97 (1H, m obsc, H_b-5′), 2.59-2.65 (1H, m obsc, H_a-4′), 2.77-2.81 (1H, m obsc, H_b-4′), 3.58-3.67 (2H, m obsc, H_a-7′ and H_b-7′), 3.77 (3H, s, H-2′), 3.81 (3H, s, H-3′), 3.84 (1H, d, J 4.5 Hz, H-12a), 4.20 (1H, d, J 12.0 Hz, H_a-6), 4.64 (1H, dd, J 3.0, 12.0 Hz, H_b-6), 4.91 (1H, ddd, J 1.0, 3.0, 4.0 Hz, H-6a), 6.46 (1H, s, H-4), 6.48 (1H, d, J 9.0 Hz, H-10), 6.81 (1H, s, H-1), 7.74 (1H, d, J 9.0 Hz, H-11); δ _C (125 MHz, CDCl₃) 16.0 (C-4′), 20.5 (C-8′), 26.5 (C-5′), 44.4 (C-12a), 56.0 (C-3′), 56.4 (C-2′), 66.5 (C-6), 69.3 (C-7′), 72.6 (C-6a), 78.2 (C-6′), 101.1 (C-4), 105.0 (C-1a), 109.4 (C-8), 110.5 (C-1), 112.1 (C-11a), 112.3 (C-10), 126.7 (C-11), 144.0 (C-2), 147.6 (C-4a), 149.6 (C-3), 160.1 (C-7a), 160.7 (C-9), 189.6 (C-12); NMR data for the minor diasteroisomer: δ _H (500 MHz, CDCl₃) 1.29 (3H, s, H-8′), 1.70-1.75 (1H, m obsc, H_a-5′), 1.80 (1H, s, OH-7′), 1.93-1.97 (1H, m obsc, H_b-5′), 2.59-2.65 (1H, m obsc, H_a-4′), 2.77-2.81 (1H, m obsc, H_b-4′), 3.58-3.67 (2H, m obsc, H_a-7′ and H_b-7′), 3.77 (3H, s, H-2′), 3.81 (3H, s, H-3′), 3.84 (1H, d, J 4.5 Hz, H-12a), 4.20 (1H, d, J 12.0 Hz, H_a-6), 4.63 (1H, dd, J 3.0, 12.0 Hz, H_b-6) 4.93 (1H, ddd, J 1.0, 3.0, 4.0 Hz, H-6a), 6.45 (1H, s, H-4), 6.48 (1H, d, J 9.0 Hz, H-10), 6.81 (1H, s, H-1), 7.73 (1H, d, J 9.0 Hz, H-11); δ _C (125 MHz, CDCl₃) 16.0 (C-4′), 20.8 (C-8′), 26.7 (C-5′), 44.4 (C-12a), 56.0 (C-3′), 56.5 (C-2′), 66.5 (C-6), 68.9 (C-7′), 72.4 (C-6a), 78.1 (C-6′), 101.0 (C-4), 105.0 (C-1a), 109.3 (C-8), 110.5 (C-1), 112.1 (C-11a), 112.2 (C-10), 126.7 (C-11), 144.0 (C-2), 147.6 (C-4a), 149.6 (C-3), 160.0 (C-7a), 160.6 (C-9), 189.6 (C-12); HRESIMS m/z 435.1398 [M + Na]⁺ (calcd for C₂₃H₂₄O₇Na, m/z 435.1414).

(6aS,12S,12aR)-12-Deoxo-12-hydroxydeguelin (29).⁵⁸ Prepared from 2 (120 mg, 0.305 mmol) by GP1 and obtained as a white solid (108 mg, 90%): [α]_D ²⁰ −79 (c 0.1, acetone); IR ν_max 3600-3300, 1612, 1586, 1509, 1462, 1444, 1216, 1191, 1152, 1094, 1083, 994, 808 cm^-1; ¹H NMR (500 MHz, acetone-d ₆) δ 1.35 (3H, s, H-7′), 1.36 (3H, s, H-8′), 3.45 (1H, app t, J 5.0 Hz, H-12a) 3.70 (3H, s, H-2′), 3.73 (3H, s, H-3′), 4.18 (1H, dd, J 3.5, 10.0 Hz, H_a-6), 4.54 (1H, d, J 3.5 Hz, OH-12), 4.59 (1H, dd, J 8.5, 10.0 Hz, H_b-6), 4.84 (1H, ddd, J 1.0, 5.0, 10.0 Hz, H-6a), 5.12 (1H, d, J 3.5, 4.0 Hz, H-12), 5.63 (1H, d, J 10.0 Hz, H-4′), 6.30 (1H, d, J 8.0 Hz, H-10), 6.36 (1H, s, H-4), 6.63 (1H, d, J 10.0 Hz, H-5′), 7.08 (1H, d, J 8.0 Hz, H-11), 7.19 (1H, s, H-1); ¹³C NMR (125 MHz, acetone-d ₆) δ 27.9 (C-7′), 28.1 (C-8′), 38.3 (C-12a), 55.8 (C-3′), 56.8 (C-2′), 66.6 (C-6), 67.6 (C-12), 71.4 (C-6a), 76.3 (C-6′), 101.3 (C-4), 109.2 (C-10), 109.9 (C-8), 111.5 (C-1a), 114.6 (C-1), 117.5 (C-4′), 117.6 (C-11a), 129.4 (C-11), 129.6 (C-5′), 144.4 (C-2), 149.2 (C-3), 150.2 (C-4a), 150.3 (C-9), 154.2 (C-7a); HRESIMS m/z 419.1450 [M + Na]⁺ (calcd for C₂₃H₂₄O₆Na, m/z 419.1465).

(6aR,12R,12aS)-12-Deoxo-12-hydroxytephrosin (30). Prepared from 8 (20.0 mg, 0.049 mmol) by GP1 and obtained as a white solid (18.5 mg, 92%): [α]_D ²⁰ −87 (c 0.1, CHCl₃); IR ν_max 3600-3300, 2924, 1614, 1585, 1508, 1445, 1266, 1217, 1197, 1151, 1114, 1056, 732 cm^-1; ¹H NMR (500 MHz, CDCl₃) 1.40 (3H, s, H-7′) δ 1.41 (3H, s, H-8′) 2.22 (1H, d, J 2.5 Hz, OH-12), 2.46 (1H, s, OH-12a), 3.82 (3H, s, H-2′), 3.82 (3H, s, H-3′), 4.34 (1H, dd, J 4.0, 10.5 Hz, H_a-6), 4.62 (1H, dd, J 4.0, 9.5 Hz, H-6a), 4.67 (1H, dd, J 9.5, 10.5 Hz, H_b-6), 4.80 (1H, app s, H-12), 5.56 (1H, d, J 10.0 Hz, H-5′), 6.39 (1H, s, H-4), 6.44 (1H, d, J 8.5 Hz, H-10), 6.63 (1H, d, J 10.0 Hz, H-4′), 7.09 (1H, d, J 8.5 Hz, H-11), 7.23 (1H, s, H-1); ¹³C NMR (125 MHz, CDCl₃) δ 28.0 (C-7′), 28.1 (C-8′), 56.0 (C-2′), 56.6 (C-3′), 64.9 (C-6), 68.5 (C-12a), 72.0 (C-12), 75.2 (C-6a), 76.2 (C-6′), 100.4 (C-4), 109.1 (C-1), 109.7 (C-8), 110.1 (C-10), 112.3 (C-1a), 113.0 (C-11a), 116.5 (C-4′), 129.3 (C-11), 129.3 (C-5′), 144.0 (C-3), 147.4 (C-7a), 149.4 (C-4a), 150.6 (C-2), 154.3 (C-9); HRESIMS m/z 435.1419 [M + Na]⁺ (calcd for C₂₃H₂₄O₇Na, m/z 435.1414).

(6aS,12aS,4′R,5′R)- and (6aS,12aS,4′S,5′S)-4′,5′-Dihydro-4′,5′-cisdihydroxydeguelin (31) were prepared by modifying a literature procedure.⁵⁸ OsO₄ (20 μL of a 2.5 wt% solution in t-BuOH, 0.002 mmol) was added to a solution of 2 (40 mg, 0.102 mmol), N-methylmorpholine N-oxide (14 mg, 0.122 mmol), and citric acid (39 mg, 0.203 mmol) in acetone (4.0 mL) and H₂O (1.0 mL). The mixture was stirred at rt for 28 h. EtOAc (10 mL) was then added followed by saturated aqueous Na₂SO₃ solution (10 mL) and the two phases were mixed vigorously for 10 min. The organic layer was separated, washed with H₂O (10 mL) and brine (10 mL), dried (MgSO₄), filtered, and concentrated. The yellow oily residue was subjected to flash chromatography (SiO₂, 1:2 hexanes/EtOAc) to give the mixture of diastereoisomers 31 as a white solid (32 mg, 78%, dr 89:11): IR ν_max 3600-3100, 1662, 1601, 1582, 1512, 1441, 1259, 1213, 1195, 1096, 1042, 817 cm^-1; NMR data for the major diastereoisomer [tentatively assigned as (6aS,12aS,4′S,5′S)-4′,5′-dihydro-4′,5′-cis-dihydroxydeguelin as the convex face of 2 is more accessible to reagents than the concave face⁵⁷]: δ _H (500 MHz, CDCl₃) 1.30 (3H, s, H-7′), 1.47 (3H, s, H-8′), 3.03 (1H, d, J 4.5 Hz, OH-4′), 3.67 (1H, d, J 2.0 Hz, OH-5′), 3.78 (3H, s, H-2′), 3.79 (1H, m obsc, H-5′), 3.81 (3H, s, H-3′), 3.90 (1H, d, J 4.5 Hz, H-12a), 4.23 (1H, d, J 12.5 Hz, H_a-6), 4.61 (1H, dd, J 3.5, 12.5 Hz, H_b-6), 5.04 (1H, dd, J 3.0, 4.5 Hz, H-4′), 5.07 (1H, ddd, J 1.0, 3.5, 4.5 Hz, H-6a), 6.45 (1H, s, H-4), 6.54 (1H, d, J 9.0 Hz, H-10), 6.78 (1H, s, H-1), 7.82 (1H, d, J 9.0 Hz, H-11); δ _C (125 MHz, CDCl₃) 22.7 (C-8′), 24.6 (C-7′), 44.4 (C-12a), 56.0 (C-3′), 56.5 (C-2′), 62.5 (C-4′), 66.4 (C-6), 70.6 (C-5′), 73.2 (C-6a), 79.1 (C-6′), 101.1 (C-4), 104.6 (C-1a), 109.6 (C-8), 110.5 (C-1), 112.6 (C-11a), 113.0 (C-10), 129.1 (C-11), 144.2 (C-2), 147.5 (C-4a), 149.7 (C-3), 160.1 (C-9), 161.2 (C-7a), 188.7 (C-12); NMR data for the minor diastereoisomer [tentatively assigned as (6aS,12aS,4′R,5′R)-4′,5′-dihydro-4′,5′-cis-dihydroxydeguelin]: δ _H (500 MHz, CDCl₃) 1.35 (3H, s, H-7′), 1.43 (3H, s, H-8′), 3.10 (1H, d, J 4.5 Hz, OH-4′), 3.73 (1H, d, J 2.0 Hz, OH-5′), 3.78 (3H, s, H-2′), 3.79 (1H, m obsc, H-5′), 3.82 (3H, s, H-3′), 3.90 (1H, d, J 4.5 Hz, H-12a), 4.23 (1H, d, J 12.5 Hz, H_a-6), 4.61 (1H, dd, J 3.5, 12.5 Hz, H_b-6), 5.00 (1H, dd, J 3.0, 4.5 Hz, H-4′), 5.07 (1H, ddd, J 1.0, 3.5, 4.5 Hz, H-6a), 6.44 (1H, s, H-4), 6.54 (1H, d, J 9.0 Hz, H-10), 6.77 (1H, s, H-1), 7.82 (1H, d, J 9.0 Hz, H-11); δ _C (125 MHz, CDCl₃) 22.6 (C-8′), 24.7 (C-7′), 44.6 (C-12a), 56.0 (C-3′), 56.5 (C-2′), 62.3 (C-4′), 66.4 (C-6), 70.7 (C-5′), 73.1 (C-6a), 79.2 (C-6′), 101.2 (C-4), 104.7 (C-1a), 109.3 (C-8), 110.3 (C-1), 112.8 (C-11a), 113.4 (C-10), 129.1 (C-11), 144.2 (C-2), 147.5 (C-4a), 149.7 (C-3), 160.1 (C-9), 161.2 (C-7a), 188.6 (C-12); HRESIMS m/z 429.1539 [M + H]⁺ (calcd for C₂₃H₂₅O₈, m/z 429.1549).

Preparation of and Kinetic Measurements on Mitochondrial Membranes

Bos taurus (bovine) heart mitochondrial membranes were prepared at 4 °C as described previously.⁷² Assays were performed in a SpectraMax 96-well plate reader at 32 °C. NADH:O₂ oxidoreduction by complexes I, III, and IV in membranes was measured using 50 μg mL⁻¹ membranes in 10 mM Tris-HCl (pH 7.4) and 250 mM sucrose using 120 μM NADH and supplemented with 1.5 μM horse heart cytochrome c (Sigma-Aldrich). The reaction was monitored using ε_{340–380(NADH)}= 4.81 mM ⁻¹ cm⁻¹. Catalysis was initiated by addition of NADH and maximal rates determined by linear regression.

OCR Measurements on Cultured Human Cells

PNT2, C4-2, and C4-2B cells were obtained from the Cancer Research UK Cambridge Research Institute and tested for mycoplasma before use. Cells were grown in RPMI 1640 (Gibco) supplemented with 10% FBS (Themo Fisher Scientific), 25 mM HEPES and 1 mM pyruvate. Seahorse cell plates were coated with poly-L-lysine hydrobromide (Sigma-Aldrich), washed with PBS and dried prior to cell seeding. 20,000 cells per well were plated into coated 96-well Seahorse plates and incubated for 16 h at 37 °C and 5% CO₂. The medium was exchanged for assay buffer containing RPMI without carbonate, supplemented with 4.5 g L^-1 glucose, 1 mM pyruvate, 32 mM NaCl, 2 mM GlutaMAX and 15 mg L^-1 Phenol Red (pH 7.4) and the cells placed in a CO₂-free incubator at 37 °C for 60 min. OCRs (oxygen consumption rates) were measured in a Seahorse extracellular flux analyser; after ~ 1 h of baseline measurements, 1 μM of rotenoids from ethanolic stocks diluted in media (as well as ethanol-only controls wells) were added and incubated for a further hour. After this, a mixture of 1 μM rotenone and 1 μM antimycin was added to inhibit mitochondrial respiration. To calculate the normalized rotenone-sensitive OCRs, the rotenone-insensitive rates (determined at the end of the experiment) were subtracted and the traces baselined to 100% before addition of rotenoids. The measured basal OCR prior to baselining for PNT2, C4-2, and C42B cells were 45, 71, and 79 pM min^-1.

Cell Growth Curves

PNT2, C4-2, and C4-2B cells were each seeded to a density of 1,000 cells per well in a 96 well plate in RPMI 1640 (Gibco) supplemented with 10% FBS (Themo Fisher Scientific), 25 mM HEPES and 1 mM pyruvate. After 24 h, 1 μM of rotenoid compounds from ethanolic stocks diluted in media (as well as ethanol-only controls) were added and growth monitored in an Essen Bioscience Incucyte until confluence was reached in the untreated control. Eight replicates were used per condition.

Cell Viability

Cell viability was tested with the acridine orange/DAPI method using a chemometec NC-3000. Cells were seeded to a density of 150,000 cells per well in 6-well plates and grown overnight. Rotenoids were added to 1 μM concentration from ethanolic stocks diluted in media, as well as ethanol controls wells and incubated for 48 hours. Cells were trypsinized using TrypLe Express (Gibco), quenched with complete media and cells pelleted by centrifugation at 400 x g for 5 mins. Cells were resuspended in PBS and mixed with staining solution containing acridine orange and DAPI (Chemometec), then loaded onto slides and measured according to the manufacturer’s instructions. Prior to trypsinization, no floating cells were visible in the wells and the viability of the control PNT2, C4-2, and C4-2B cells were 89, 87, and 83% respectively.

Statistical Methods

Data are presented as mean averages. IC₅₀ values were calculated in Prism version 7.0e from normalized kinetic data using a log(inhibitor) vs. normalized response fit with variable slope. Error is presented as either a 95% confidence interval or standard error of the mean and statistical significance between samples were tested in Prism via an unpaired, two-tailed t-test. * P ≤ 0.1, ** P ≤ 0.01, *** P ≤ 0.001,**** P ≤ 0.000.1.