Modular Synthesis and Biological Investigation of 5-Hydroxymethyl Dibenzyl Butyrolactones and Related Lignans

Dibenzyl butyrolactone lignans are well known for their excellent biological properties, particularly for their notable anti-proliferative activities. Herein we report a novel, efficient, convergent synthesis of dibenzyl butyrolactone lignans utilizing the acyl-Claisen rearrangement to stereoselectively prepare a key intermediate. The reported synthetic route enables the modification of these lignans to give rise to 5-hydroxymethyl derivatives of these lignans. The biological activities of these analogues were assessed, with derivatives showing an excellent cytotoxic profile which resulted in programmed cell death of Jurkat T-leukemia cells with less than 2% of the incubated cells entering a necrotic cell death pathway.

Owing to their anti-cancer properties and their classification as drug-like compounds [13] extensive work has gone into the study of these compounds and their related analogues to explore and establish structure-activity relationships and the possible use of these lignans as lead compounds for therapeutics. Whilst previous work has explored the synthesis of these lignans and analogues thereof [14][15][16], mainly focusing on changing the substituents on the aryl rings [17], one area that has not been extensively investigated is the synthesis of C-5 substituted analogues of these butyrolactone lignans, represented by 4. We have previously shown that the acyl-Claisen rearrangement can be used to prepare disubstituted morpholine pentenamides 5 with high diastereoselectivity at the C-3 and C-4 positions which correspond to the benzyl groups in the lactone scaffold ( Figure 2) [18][19][20][21][22]. Furthermore, in our efforts to a prepare a number of different lignan scaffolds [18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36], we have used amides such as 5 to prepare compounds including tetrahydrofuran lignans (e.g., galbelgin 6), aryltetralins (e.g., ovafolinin 7) and aryl dihydronaphthalene lignans (e.g., (-)-pycananthuligene B 8). We wished to explore the usage of this methodology to synthesise butyrolactone lignans, as well as probe the effect of adding a substituent at the C-5 position on the biological activity. The route would be convergent and modular, allowing for simple modification of aromatic groups resulting in the synthesis of a number of analogues. 2

. Results and Discussion
In order to utilise the acyl-Claisen rearrangement to prepare the desired lactones, the corresponding allylic morpholines and acid chlorides first needed to be synthesised. Allylic morpholines 9a and 9b were synthesised in five steps from 4-allyl-1,2-dimethoxybenzene 10 and We wished to explore the usage of this methodology to synthesise butyrolactone lignans, as well as probe the effect of adding a substituent at the C-5 position on the biological activity. The route would be convergent and modular, allowing for simple modification of aromatic groups resulting in the synthesis of a number of analogues.

Results and Discussion
In order to utilise the acyl-Claisen rearrangement to prepare the desired lactones, the corresponding allylic morpholines and acid chlorides first needed to be synthesised. Allylic morpholines 9a and 9b were synthesised in five steps from 4-allyl-1,2-dimethoxybenzene 10 and safrole 11 (Scheme 1), respectively. Firstly, allylic benzenes 10 and 11 were dihydroxylated using catalytic osmium tetroxide giving 12 and 13, followed by periodate cleavage to give aldehydes 14 and 15. Aldehydes 14 and 15 were immediately used in a Wittig reaction with (carbethoxymethylene)-triphenylphosphorane to exclusively give the E-isomer of α,β-unsaturated esters 16 and 17, in 55% and 56% yields, respectively, over three steps. The esters 16 and 17 were then reduced to allylic alcohols 18 and 19 using di-iso-butyl aluminium hydride (DIBAL-H) in excellent yields. Alcohols 18 and 19 were then converted to the corresponding allylic morpholines 9a and 9b, by first generating a mesylate in situ, which then underwent substitution to give allylic morpholines 9a and 9b. The required acid chlorides were then synthesised in four or five steps from commercially available benzaldehydes-piperonal 20, 3,4,5-trimethoxybenzaldehyde 21 and vanillin 22 (Scheme 2).
Acyl-Claisen rearrangements were undertaken using two allylic morpholines 9a and 9b which were reacted individually with the four acid chlorides 34a-d, using TiCl 4 ·2THF as the Lewis acid, providing eight morpholine amides 35aa-bd in 42-95% yields. All amides 35aa-bd were obtained as single diastereomers with a syn-configuration between the C-2 and C-3 substituents (Scheme 3). In all cases it was observed that only the 3,4-trans-4,5-trans-lactone was obtained. This configuration was confirmed through NOESY NMR analysis, depicted in Figure 3 with 4bb. We propose that only this isomer was obtained due to the preferential cyclisation of the 3,4-anti diol 36, leaving the polar uncyclised 3,4-syn diols 37 which were difficult to isolate. Upon dihydroxylation of amide 35bb at a larger scale and following isolation of lactone 4bb by column chromatography, a small sample of the corresponding uncyclised diol 37 was able to be isolated. This diol 37 was subsequently cyclised using 2 M H 2 SO 4 in methanol to give the corresponding C-5 epimer, epi-4bb, confirming this hypothesis (Scheme 4). Acyl-Claisen rearrangements were undertaken using two allylic morpholines 9a and 9b which were reacted individually with the four acid chlorides 34a-d, using TiCl4·2THF as the Lewis acid, providing eight morpholine amides 35aa-bd in 42-95% yields. All amides 35aa-bd were obtained as single diastereomers with a syn-configuration between the C-2 and C-3 substituents (Scheme 3). All amides 35aa-bd then underwent dihydroxylation using osmium tetroxide and N-methylporpholine N-oxide (NMO) to give cyclized 5-hydroxymethyllactones 4aa-bd.
In all cases it was observed that only the 3,4-trans-4,5-trans-lactone was obtained. This configuration was confirmed through NOESY NMR analysis, depicted in Figure 3 with 4bb. We propose that only this isomer was obtained due to the preferential cyclisation of the 3,4-anti diol 36, leaving the polar uncyclised 3,4-syn diols 37 which were difficult to isolate. Upon dihydroxylation of amide 35bb at a larger scale and following isolation of lactone 4bb by column chromatography, a small sample of the corresponding uncyclised diol 37 was able to be isolated. This diol 37 was subsequently cyclised using 2 M H2SO4 in methanol to give the corresponding C-5 epimer, epi-4bb, confirming this hypothesis (Scheme 4).   Finally, to deprotect the benzyl-protected lactones 4ad and 4bd to their respective alcohols, they were subjected to hydrogenolysis to give 4ae and 4be in excellent yields. Transformation of C-5 hydroxymethyl analogues 4 into dibenzylbutryolactone lignans 1 was achieved via reduction using LiAlH 4 , to the corresponding triols 38aa-bd, followed by periodate cleavage, forming lactols 39aa-bd. These lactols 39aa-bd were then oxidised using Fetizon's reagent [37,38] to give racemic samples of dibenzyl butyrolactone lignans 1aa-bd, including known natural products arcitin 1aa, bursehernin 1ab, (3R*,4R*)-3-(3 ,4 -dimethoxybenzyl)-4-(3 ,4 ,5 -trimethoxybenzyl)dihydrofuran-2(3H)-one 1ac, kusunokinin 1ba, hinokinin 1bb, and isoyatein 1bc. Additionally, phenolic lignans, buplerol 1ae, and haplomyrfolin 1be were produced by the debenzylation of 1ad and 1bd, respectively. Several of the synthesised compounds were then tested for their anti-microbial and cytotoxic activities. All tested compounds were found to be inactive against Staphlycoccus aureus and Escherichia. coli, showing no to little antimicrobial activity, while the compounds were shown to exhibit antiproliferative effects against Jurkat T-leukaemia cells, while also showing effects on cell cycle progression (Figure 4). While the synthesised naturally-occurring dibenzyl butyrolactones, arcitin 1aa, bursehernin 1ab, and (3R*,4R*)-3-(3′′,4′′-dimethoxybenzyl)-4-(3′,4′,5′trimethoxybenzyl)dihydrofuran-2(3H)-one 1ac, boasted the best activities, 5-hydroxymethyl analogue 4bb had similar potency. Compound 4bb was shown to have the best activity of all of the 5-hydroxymethyl analogues tested, inducing apoptosis, evidenced by the presence of cells in the early and predominantly in the late apoptotic cell cycle ( Figure 4). Additionally the compounds demonstrated an effect on cell cycle progression. A significantly greater number of 4N cells were present following treatment with compound 4bb in particular causing a significant increase in 4N cells ( Figure 4D and 4E). During the cell cycle, DNA is replicated in the S-phase, going from 2N in G1, to 4N by the end of this phase. The DNA content in cells then remains at 4N during G2 and M phases, before cytokinesis at the M-phase. The observation that there was in increase in 4N cells indicates that it is likely these cells have arrested in G2/M and will not re-enter next G1-phase after this mitotic slippage. This is in-line with published cell cycle data following treatment with other lignans [39,40]. Furthermore, our compounds showed minimal levels of necrosis, less than 2% (except 4ba with 7%), suggesting that the cells are in fact entering programmed cell death cycles, which is considered the most effective and non-inflammatory mechanism of cancer-cell death.
In conclusion, the synthesis of dibenzyl butyrolactone lignans utilising the acyl-Claisen rearrangement has been accomplished and represent a new, modular, and convergent method towards the synthesis of this class of natural products. Furthermore, this route gives rise to the previously-unexplored 5-hydroxymethyl derivatives 4 of these natural products. The biological activities of this new set of derivatives were assessed, with one derivative in particular, 4bb, showing a superior cytotoxic profile and resulting in cell cycle arrest and programmed cell death of Jurkat Tleukaemia cells with less than 2% of the incubated cells entering a necrotic cell death pathway. Several of the synthesised compounds were then tested for their anti-microbial and cytotoxic activities. All tested compounds were found to be inactive against Staphlycoccus aureus and Escherichia. coli, showing no to little antimicrobial activity, while the compounds were shown to exhibit antiproliferative effects against Jurkat T-leukaemia cells, while also showing effects on cell cycle progression ( Figure 4). While the synthesised naturally-occurring dibenzyl butyrolactones, arcitin 1aa, bursehernin 1ab, and (3R*,4R*)-3-(3 ,4 -dimethoxybenzyl)-4-(3 ,4 ,5 -trimethoxybenzyl)dihydrofuran-2(3H)-one 1ac, boasted the best activities, 5-hydroxymethyl analogue 4bb had similar potency. Compound 4bb was shown to have the best activity of all of the 5-hydroxymethyl analogues tested, inducing apoptosis, evidenced by the presence of cells in the early and predominantly in the late apoptotic cell cycle ( Figure 4). Additionally the compounds demonstrated an effect on cell cycle progression. A significantly greater number of 4N cells were present following treatment with compound 4bb in particular causing a significant increase in 4N cells ( Figure 4D,E). During the cell cycle, DNA is replicated in the S-phase, going from 2N in G 1 , to 4N by the end of this phase. The DNA content in cells then remains at 4N during G 2 and M phases, before cytokinesis at the M-phase. The observation that there was in increase in 4N cells indicates that it is likely these cells have arrested in G2/M and will not re-enter next G 1 -phase after this mitotic slippage. This is in-line with published cell cycle data following treatment with other lignans [39,40]. Furthermore, our compounds showed minimal levels of necrosis, less than 2% (except 4ba with 7%), suggesting that the cells are in fact entering programmed cell death cycles, which is considered the most effective and non-inflammatory mechanism of cancer-cell death.
In conclusion, the synthesis of dibenzyl butyrolactone lignans utilising the acyl-Claisen rearrangement has been accomplished and represent a new, modular, and convergent method towards the synthesis of this class of natural products. Furthermore, this route gives rise to the previously-unexplored 5-hydroxymethyl derivatives 4 of these natural products. The biological activities of this new set of derivatives were assessed, with one derivative in particular, 4bb, showing a superior cytotoxic profile and resulting in cell cycle arrest and programmed cell death of Jurkat T-leukaemia cells with less than 2% of the incubated cells entering a necrotic cell death pathway.

General Methods
All reactions were carried out with oven-dried glassware and under a nitrogen atmosphere in dry, freshly distilled solvents unless otherwise noted. Diisopropylethylamine was distilled from CaH 2 and stored over activated 4Å molecular sieves. All melting points for solid compounds, given in degrees Celsius ( • C), were measured using a Reicher-Kofler block and are uncorrected. Infrared (IR) spectra were recorded using a Perkin Elmer Spectrum1000 FT-IR spectrometer. The NMR spectra were recorded on a 400 MHz spectrometer. Chemical shifts are reported relative to the solvent peak of chloroform (δ 7.26 for 1 H and δ 77. 16 ± 0.06 for 13 C). The 1 H-NMR data was reported as position (δ), relative integral, multiplicity (s, singlet; d, doublet; dd, doublet of doublets; ddd, doublet of doublet of doublets; dt, doublet of triplets; dq, doublet of quartets; t, triplet; td, triplet of doublets; q, quartet; m, multiplet), coupling constant (J, Hz), and the assignment of the atom. The 13 C-NMR data were reported as position (δ) and assignment of the atom. The NMR assignments were performed using COSY, HSQC and HMBC experiments. High-resolution mass spectroscopy (HRMS) was carried out by electrospray ionization (ESI) on a MicroTOF-Q mass spectrometer. Fetizon's reagent was prepared following a literature procedure [41]. Unless noted, chemical reagents were used as purchased.

General Procedure A: Acyl-Claisen
To a stirred suspension of TiCl 4 ·2THF (1 mmol) in CH 2 Cl 2 (5 mL), under an atmosphere of nitrogen, was added a solution of allylic morpholine (1 mmol) in CH 2 Cl 2 (2.5 mL) followed by dropwise addition of i Pr 2 NEt (1.5 mmol). After stirring for 10 min a solution of acid chloride (1.2 mmol) in CH 2 Cl 2 (2.5 mL) was added dropwise and the resultant mixture stirred for the specified time. The reaction mixture was quenched with aqueous NaOH (12 mL, 1 M) and the aqueous phase extracted with CH 2 Cl 2 (3 × 10 mL). The combined organic extracts were washed with brine (6 mL), dried (MgSO 4 ), the solvent removed in vacuo and the crude product purified by column chromatography.

General Procedure B: Dihydroxylation
To a stirred solution of morpholine pentenamide (1 mmol) in t BuOH/H 2 O (1:1, 20 mL) or t BuOH/H 2 O/THF (1:1:1, 30 mL) was added NMO (3 mmol). A solution of OsO 4 (0.08 mmol, 2.5% w/v in t BuOH) was then added dropwise and the resultant mixture stirred for the specified time. The mixture was quenched with saturated aqueous Na 2 SO 3 (30 mL) and stirred for a further 1 h. The aqueous phase was extracted with ethyl acetate (3 × 20 mL), the combined organic extracts washed with aqueous KOH (5 mL, 1 M), dried (MgSO 4 ), the solvent removed in vacuo and the crude product purified by column chromatography.

General Procedure C: Lithium Aluminum Hydride Reduction
To a stirred suspension of LiAlH 4 (1.4 mmol) in THF (10 mL), under an atmosphere of nitrogen at 0 • C, was added a solution of lactone (1 mmol) in THF (10 mL) and the mixture stirred for the specified time. After warming to room temperature, the mixture was quenched with the addition of water (30 mL) and the aqueous phase extracted with ethyl acetate (3 × 40 mL). The combined organic extracts were washed with brine (25 mL), dried (MgSO 4 ), and the solvent removed in vacuo.

General Procedure D: Periodate Cleavage
To a stirred solution of triol (1 mmol) in MeOH/H 2 O (3:1, 50 mL) was added NaIO 4 (1.2 mmol) and the resultant mixture stirred for the specified time. The reaction mixture was quenched with brine (40 mL) and extracted with ethyl acetate (3 × 80 mL). The organic layers were combined, washed with water (2 × 40 mL), dried (MgSO 4 ), and solvent removed in vacuo to give the crude product which was purified by column chromatography if necessary.

General Procedure E: Fétizon's Oxidation
To a stirred solution of lactol (1 mmol) in toluene (60 mL), under an atmosphere of nitrogen, was added Fétizon s reagent (2 mmol) and heated at reflux for the specified time. The reaction mixture was allowed to cool and filtered, the solvent removed in vacuo and the crude product purified by column chromatography.

General Procedure F: Benzyl Deprotection
To a stirred solution of benzyl ether (1 mmol) in MeOH (30 mL) was added 10% palladium on carbon (20% w/w) and the resultant mixture stirred under and atmosphere of hydrogen for the specified time. The reaction mixture was filtered through celite, washed with methanol (3 × 20 mL), the solvent removed in vacuo and the crude product purified by column chromatography if necessary (The 1 H and 13 C-NMR spectra of compounds in the Supplemental Materials).
To a stirred solution of unsaturated ester 25 (4.13 g, 18.6 mmol) in ethyl acetate (30 mL) was added 10% palladium on activated carbon (0.4 g, 10% w/w). The solution was flushed with an atmosphere of hydrogen and stirred for 2 h. The reaction mixture was then filtered through a plug of celite and washed with ethyl acetate, solvent was then removed in vacuo to give saturated ester 28 (3.9 g, 94%) as a yellow oil which was then used without further purification. To a stirred solution of phenol 28 (3.75 g, 16.7 mmol) in acetonitrile (40 mL), under an atmosphere of nitrogen, was added K 2 CO 3 (6.9 g, 50.0 mmol) and stirred for 10 min. Benzyl bromide (6.0 mL, 50.0 mmol) was then added and the resulting mixture allowed to stir for 65 h. The reaction mixture was then quenched with addition of water (50 mL) and extracted with CH 2 Cl 2 (3 × 30 mL). The organic phases were combined, washed with water (2 × 10 mL) and dried (MgSO 4 ). Solvent was then removed in vacuo and the crude product purified by column chromatography (9:1 hexanes, ethyl acetate) to give benzyl ether 29 (4.38 g, 83%) as a colourless oil which was used immediately. To a stirred solution of ester 29 (4.3 g, 13.7 mmol) in methanol (30 mL) was added aqueous NaOH (55 mL, 1 M, 4 eq.) and stirred for 2.

3-(3 ,4 -Methylenedioxyphenyl)propanoyl chloride (34b).
To a stirred solution of carboxylic acid 30 (0.22 g, 1.2 mmol) in CH 2 Cl 2 (3 mL), under an atmosphere of nitrogen, was added oxalyl chloride (0.2 mL, 2.3 mmol) dropwise and the mixture stirred for 4 h. The solvent was removed in vacuo to give the title compound 34b (0.24 g, quant.) as a green oil, which was placed under nitrogen and used without further purification.
3-(3 ,4 -Dimethoxyphenyl)propanoyl chloride (34a). To a stirred solution of carboxylic acid 33 (0.24 g, 1.2 mmol) in CH 2 Cl 2 (5 mL), under an atmosphere of nitrogen, was added oxalyl chloride (0.2 mL, 2.3 mmol) dropwise and the mixture stirred for 2.5 h. The solvent was removed in vacuo to give the title compound 34a (0.26 g, quant.) as a yellow oil, which was placed under nitrogen and used without further purification.

Annexin V/PI Assay
Following treatments at a cell density of 1 × 10 5 cells/well, the samples were centrifuged at 500 g for 5 min and the supernatant was removed. Cells were washed in 500 µL DPBS before addition of 100 µL of 1 × Annexin V binding buffer (BD Biosciences). A 5 µL volume of FITC-conjugated Annexin V (BD Biosciences) and 10 µL Propidium Iodide (BD Biosciences) was added and the cells were incubated in the dark for 20 min. Samples were diluted by addition of 400 µL 1 × Annexin V binding buffer before immediate analysis on an Accuri C6 Flow Cytometer (Becton Dickinson, Oxford, UK).

Cell Cycle Analysis
Following treatments at a cell density of 5 × 10 6 /well, cells were centrifuged at 500 g for 5 min and the supernatant was removed. The remaining cell pellet was vortexed while simultaneously adding 500 µL of 70% ethanol dropwise, fixing the cells and minimising clumping. The samples were incubated at 4 • C for 30 min, and then centrifuged at 1000 g for 5 min. The supernatant was discarded, and the pellet was re-suspended in 500 µL DPBS. The samples were centrifuged again at 1000 g for 5 min, and the supernatant was removed a final time. The pellet was resuspended in 50 µL RNase A (100 µg/mL stock; Roche, UK) and 200 µL PI (50 µg/mL stock; Sigma, UK). The samples were analyzed on an Accuri C6 flow cytometer (Becton Dickinson) and data was modelled and interpreted using ModFit Analysis Software, version 5.0 (Verity Software House).