Pseudopteroquinoxaline (5) was synthesized in one-pot by oxidation of the pseudopterosin G–J aglycone (4) [18] with Ag₂O and condensation with ethylenediamine (Scheme 1). In an alternative synthesis, treatment of 4 with Dess-Martin periodinane in DCM/MeOH, followed by reaction with ethylenediamine yielded 5 and the tertiary ether 6 as a minor side product.

Synthesis of the novel derivatives 21-((1H-imidazol-5-yl)methyl)-pseudopteroxazole (7a) and the regioisomer (7b) followed the previously reported general method utilizing the aglycone (4), Ag₂O and histidine (Scheme 2). After purification by flash chromatography, a mixture of 7a and 7b was obtained in a 2.4:1 ratio as determined by ¹H NMR analysis. This product regioisomer ratio differs from the ~10:1 ratio previously observed with other amino acids [11]. While the reasons behind this are under further investigation, it is conceivable that the nucleophilic imidazole attacks the ortho-quinone at C-10, which increases the relative rate of condensation at C-9. Separation of the regioisomers (7a/7b) proved challenging; while HPLC did not lead to peak resolution, a portion of the material was slightly enriched in 7a (3:1 ratio) by peak shaving. This material (7a/7b) has been thoroughly and unambiguously characterized; spectra and analytical chromatograms are provided in the supporting information.

The mono-pentyl ether (8) derivative of the pseudopterosin G–J aglycone (4) was synthesized to provide a phenolic mimic of homopseudopteroxazole (2), given that 2 possesses a pentyl chain and was reportedly active against M. tuberculosis H₃₇Rv [7]. The mono-pentyl ether (8) and mono-methyl ether (9) [19] were synthesized by alkylation of 3a–d with iodopentane or iodomethane, respectively, followed by acid catalyzed hydrolysis of the fucose moiety (Scheme 3). Further substitution of the free phenol in 9 by treatment with the appropriate electrophile yielded the di-methyl ether (10), the triflate (11) and the carbamate (12).

The syntheses of the prenylated aromatic mimics of pseudopterosin are shown in Scheme 4. Acid catalyzed reaction of 2,6-dimethoxyphenol (13) with 2-methyl-3-buten-2-ol yielded the mono-, di- and tri-prenylated derivatives (14, 15 & 16). Compound 13 was further utilized as a model compound to develop conditions suitable for the glycosylation of these phenols. Treatment of 13 with the benzoylated glycosyl donor 21 and BF₃·Et₂O yielded 17, which was deprotected with K₂CO₃ to give the galactoside 18 in high yield. Identical reaction sequences utilizing 14 gave the benzoylated glycoside 19 followed by the desired prenylated galactoside 20. Attempts to glycosylate 15 were unsuccessful.

The biological activities of fifteen semi-synthetic pseudopteroxazoles and isopseudopteroxazoles are shown in Table 1. The minimum inhibitory concentrations (MICs) against M. tuberculosis H₃₇Rv (ATCC 27294) were determined in vitrousing the microplate Alamar blue assay (MABA) [20]. Generally substitutions at the C-21 oxazole moiety in 1 lead to congeners with reduced activity against M. tuberculosis H₃₇Rv, however, three compounds (7a/7b, 22 & 25) showed activity against M. tuberculosis H₃₇Rv comparable to 1. Semi-synthetic homopseudopteroxazole (2) was not active against M. tuberculosis H₃₇Rv in contrast to the literature report for natural 2 isolated from P. elisabethae, which was reported to show 80% inhibition against M. tuberculosis H₃₇Rv at 12.5 μg/mL (40 μM) [7]. Our result with 2 was consistent with the inactivity of other members of the series with lipophilic C-21 substituents.

As a measure of potential toxicity the IC₅₀ values of the pseudopteroxazoles against Vero cells [21] were determined, and the selectivity index was calculated (Vero cell IC₅₀/M. tuberculosis H₃₇Rv MIC). Only four semi-synthetic pseudopteroxazoles were non-toxic to Vero cells at 128 μg/mL: pseudopteroxazole (1), homopseudopteroxazole (2), the methionine derivative (27) and the histidine derivative (7a/7b, 4% toxicity at 128 μg/mL). Of these, two (1 and 7a/7b) exhibited antitubercular activity resulting in selectivity indices >8.6. The cytotoxicity result for 1 is comparable to that observed with “natural” pseudopteroxazole (1), which was reported to show no significant cytotoxicity against the NCI-60 cell line assay [6]. Interestingly, isopseudopteroxazole (22) displayed toxicity towards the Vero cells (IC₅₀ 52 μg/mL), otherwise this regioisomer exhibited very similar antimicrobial activity to 1.

The pseudopteroxazoles were also tested at a single point concentration (64 μg/mL) in the low-oxygen-recovery assay (LORA), a model of NRP-TB [22]. Selected active compounds were further tested to determine LORA MICs; 7a/7b showed the strongest activity, with an MIC of 12 μg/mL (Table 1). As antibiotics acting on the cell wall are typically not active against NRP-TB, the LORA activity exhibited by pseudopteroxazole and several semi-synthetic congeners suggests that the target of these compounds is not the cell wall [22].

The semi-synthetic histidine derivative (7a/7b) is the most promising of the pseudopteroxazoles as it was non-toxic, exhibited potent broad-spectrum Gram-positive antibiotic activity and was the most active compound in assays against replicating and NRP-TB (MABA and LORA, respectively). While the MIC of 7a/7b against M. tuberculosis H₃₇Rv was moderate (13 μg/mL or 34 μM) in comparison to the first line drugs isoniazid and rifampin, its activity compares favorably to other first and second line TB treatments such as ethambutol (4.6–9.2 μM), kanamycin (2.5–10.3 μM), capreomycin (0.94–3.7 μM) and cycloserine (122–490 μM) [23]. One of the most promising attributes of 7a/7b was that it showed greater potency in LORA (LORA MIC 12 μg/mL or 31 μM) compared to 1 (LORA MIC 50 μg/mL or 162 μM). Activity against NRP-TB is highly desirable given this phenotype contributes to lengthy treatment regimens leading to poor patient compliance ultimately translating into increased TB transmission rates and selection for drug-resistant strains [4]. As shortening of treatment times is a key goal of current TB chemotherapeutic research, this compound may represent a starting point for developing drugs that are more efficacious towards latent TB infections [1].

Recently a diterpene that is structurally related to pseudopterosins has been shown to possess activity against M. tuberculosis H₃₇Rv and a multidrug-resistant strain [24]. These examples suggest that the semi-synthetic pseudopteroxazole congeners may also display activity against drug-resistant strains. Thus, we determined MICs of 1 against six isogenic mono-resistant M. tuberculosis H₃₇Rv strains (Table 2). In our study the six strains tested were singly-resistant to a structurally diverse group of antibiotics targeting a variety of cellular targets. The aminoglycosides streptomycin and kanamycin target the 30S ribosome, the fluoroquinolone moxifloxacin targets DNA gyrase, the ansamycin antibiotic rifampin targets RNA polymerase and the small heterocyclic antibiotics isoniazid and cycloserine inhibit cell wall biosynthesis, albeit via distinct mechanisms [25,26,27]. Pseudopteroxazole exhibited virtually identical activity against wild-type and antibiotic resistant strains. While 1 was significantly less potent than rifampin and isoniazid none of the antibiotic resistant strains exhibited cross-resistance to 1, suggesting that it may exert its antimicrobial activity via a unique mode of action.

The biological data of the pseudopterosin derivatives (including the pseudopteroquinoxalines) and of the prenylated aromatic mimics of pseudopterosins are summarized in Table 3. These compounds were tested in vitro against MRSA, VRE, M. smegmatis (ATCC 12051), and M. diernhoferi (ATCC 19340) using microbroth dilution antibiotic susceptibility assays. The compounds were also assessed for activity against M. tuberculosis, Vero cells and in the LORA. Natural pseudopterosins G–J (3a–d) showed the strongest activity against all pathogens, and exhibited low toxicity towards Vero cells. The mono-methyl ether (9) retained some activity against all pathogens; however, all other semi-synthetic derivatives showed significantly reduced activity against one or more organisms. Pseudopteroquinoxaline (5) was moderately active against the three mycobacteria; however, it was toxic towards the Vero cells.

The activity of the prenylated aromatic mimics is interesting: the mono-prenylated compound (14) was only weakly active against M. smegmatis and M. diernhoferi; the tri-prenylated compound (16) showed moderate activity against M. diernhoferi, and was also active against VRE and M. tuberculosis; the di-prenylated compound (15) was the most active as it showed good to moderate activity against all bacteria with an IC₅₀ of 3 μg/mL against VRE. To provide a glycoside mimic of pseudopterosins, the galactoside derivative 20 was synthesized from 14. This synthetic galactoside was less active than the parent prenylated mimic and no additional glycosides were synthesized following unsuccessful glycosylation attempts utilizing the di-prenylated compound (15). The activity of the prenylated aromatics hints at the possibility of a simpler pharmacophore than the natural diterpene skeleton. However, more work is required here as it does not necessarily follow that these mimics operate through the same mechanism of action as the pseudopterosins/pseudopteroxazoles.

Due to the challenges associated with cultivating M. tuberculosis (slow growth rate, biosafety risk, biosafety level 3 containment requirements) we initially used two fast growing mycobacteria, M. smegmatis and M. diernhoferi, as model organisms to evaluate the antimycobacterial activity of the compounds described herein. The usefulness of fast growing mycobacteria to detect compounds inhibitory to the growth of TB, particularly the widely used M. smegmatis, has recently been questioned. In a comprehensive study of the relative activity of 5000 compounds against M. smegmatis and M. tuberculosis, 50% of compounds active against M. tuberculosis were not detected as active against M. smegmatis [27]. Despite this disparity, the use of fast growing mycobacterial models continue to have utility as a whole cell screen against M. smegmatis identified the promising new diarylquinoline, TMC207, which is currently in Phase II–III clinical trials for the treatment of multidrug-resistant TB [1]. In the evaluation of the series of pseudopteroxazole congeners the model mycobacteria were good predictors of antimycobacterial activity (Table 1 and Table 3, and previously published data [11]). There were no instances of false-negative predictions (i.e., the model organisms were insensitive to a compound which inhibited the growth of M. tuberculosis). However, there were a few instances where the model organisms predicted activity which was not mirrored by M. tuberculosis in the MABA (4, 27, 28, 31). Interestingly, in three of these cases significant activity was observed in the LORA. These observations suggest that model mycobacteria can be a reliable predictor of M. tuberculosis activity for a particular series of molecules.

A sample of pseudopterosin G–J (3a–d, 20 mg, 0.041 mmol) was refluxed in methanolic HCl (1.5 N, 10 mL) under N₂ for 2.5 h. The crude mixture was then partitioned between EtOAc and H₂O and the EtOAc phase was concentrated in vacuo to give the crude aglycone (4). This material was dissolved in EtOH (15 mL) and air was bubble through the sample for 10 min. Ethylenediamine (50 μL, excess) and Ag₂O (14 mg, 0.065 mmol) were then added and the reaction was refluxed for 1.5 h. The reaction mixture was then filtered through Celite and partitioned between EtOAc and H₂O. The EtOAc phase was subjected to the standard work up procedure and then purified by flash chromatography (Silica, hexane→MTBE gradient) to give 5 (2 mg, 0.0063 mmol, 15% over two steps).

5: yellow oil; [α]²⁵_D +103 (c 0.03, CHCl₃); IR ν_max 2921, 2858, 1470 cm⁻¹; ¹H NMR (CDCl₃, 600 MHz) δ 8.74 (m, 2H, H-20, H-21), 5.04 (d, 1H, J = 9.1 Hz, H-14), 4.11 (app. q, 1H, J = 8.5 Hz, H-1), 4.06 (app. q, 1H, J = 7.3 Hz, H-7), 2.64 (s, 3H, H-20), 2.39-2.33 (m, 1H), 2.28–2.23 (m, 1H), 2.23–2.18 (m, 1H), 2.17–2.11 (m, 1H), 1.82 (s, 3H, H-17), 1.70 (s, 3H, H-16), 1.54 (m, 1H), 1.42–1.40 (m, 2H), 1.39 (d, 3H, J = 6.9 Hz, H-19), 1.37–1.32 (m, 1H), 1.26–1.24 (m, 1H), 1.11 (d, 3H, J = 6.2 Hz, H-18) 1.14–1.09 (m, 1H); ¹³C NMR (CDCl₃, 150 MHz) δ 142.1, 142.0, 141.9, 141.3, 141.0, 140.1, 137.0, 132.2, 129.7, 129.7, 44.3, 39.5, 37.4, 33.8, 30.7, 28.5, 26.8, 25.5, 24.7, 20.4, 17.7, 12.8; APCIMS m/z 321 [M + H]⁺; HRMS-ES m/z [M + H]⁺321.2313 (calcd for C₂₂H₂₉N₂, 321.2325).

To a solution of the pseudopterosin aglycone (4, 26.3 mg, 0.088 mmol) in DCM (10 mL + 100 μL H₂O) was added Dess-Martin periodinane (68 mg, 0.16 mmol). After stirring for 15 min, MeOH (2 mL) and ethylenediamine (1 mL) were added. After another 45 min the solvent was removed in vacuo, and then isopropyl alcohol was added (20 mL). After refluxing overnight additional ethylenediamine (200 μL) was added and the solution was refluxed for a further 24 h. The reaction products were partitioned between EtOAc and H₂O and the organic phase was subjected to the standard work up procedure to give an orange brown gum (31.8 mg). Purification by flash chromatography (diol modified silica, hexane→MTBE gradient) yielded the quinoxaline (5, 4.6 mg, 16%) and the methyl ether (6, 1.6 mg, 5%).

6: amorphous semi solid; [α]²⁵_D −5.0 (c 0.08, CHCl₃); IR ν_max 2925, 2866, 1470, 1079 cm⁻¹; ¹H NMR (CDCl₃, 600 MHz) δ 8.76 (s, 2H, H-20, H-21), 3.89 (app. q, 1H, J = 6.9 Hz), 3.71 (m, 1H), 3.20 (s, 3H, OMe), 2.79 (s, 3H, H-20), 2.49 (m, 1H), 2.21–2.15 (m, 3H), 1.89 (dd, 1H, J = 9.5, 14.5 Hz, H-14-a), 1.68–1.62 (m, 2H), 1.54–1.50 (m, 1H), 1.41 (d, 3H, J = 6.9 Hz, H-19), 1.27 (s, 6H), 1.26 (m, 2H), 1.13 (d, 3H, J = 6.5 Hz); ¹³C NMR (CDCl₃, 150 MHz) δ 144.7, 142.3, 141.8, 141.6, 141.3, 140.4, 136.5, 129.8, 75.3 (C-15), 49.2 (C-23), 48.6 (C-14), 42.3, 38.2, 34.7, 32.2, 30.3, 29.4, 25.8, 25.7, 25.4, 24.0, 21.0. 12.6; APCIMS m/z 353 [M + H]⁺; HRMS-ES m/z [M + H]⁺353.2582 (calcd for C₂₃H₃₃N₂O, 353.2587).

The pseudopteroxazole C-21 (1H-imidazol-4-yl)methyl derivatives (7a and 7b) were synthesized from the pseudopterosin aglycone (4, 193 mg, 0.64 mmol), Ag₂CO₃ (1.4 equiv.) and histidine (6.7 equiv.) following the previously reported general procedure [11]. After purification by flash chromatography (diol modified silica, hexane→MTBE gradient) the product was obtained in 23% isolated yield (57.5 mg, 2.4:1 ratio of 7a/7b). A portion of this material was subjected to RP-HPLC (Phenomenex, phenylhexyl, 5 μm, 250 × 10 mm, 2.9 mL/min) eluted with MeOH:H₂O:HCO₂H (70:30:0.1). While the regioisomers eluted as one asymmetric peak (19.8 to 20.7 min), peak shaving lead to the isolation of an enriched fraction (3:1 ratio of normal to inverse regioisomer), which was the material used for biological evaluation.

7a/7b (3:1 ratio): orange immobile oil; [α]²⁵_D +129.4 (c 0.09, CHCl₃); IR ν_max 2948, 2921, 2856, 1446, 1085 cm⁻¹; ¹H and ¹³C NMR see Supplementary Table S1. APCIMS m/z 390 [M + H]⁺; MSMS spectrum see Supplementary Figure S9; HRMS-ES m/z [M + H]⁺ 390.2549 (calcd for C₂₅H₃₂N₃O, 390.2540).

A solution of pseudopterosins G–J (3a–d, 102 mg, 0.2 mmol), K₂CO₃ (1 g, excess) and iodopentane (2 mL, excess) in anhydrous acetone was refluxed under N₂ for 18 h. The products were then partitioned between EtOAc and H₂O and subjected to the standard work up procedure to give an orange/brown oil (127 mg). This crude product was then dissolved in 1.5 M HCl in MeOH (10 mL) and refluxed under N₂ for 2 h. The products were partitioned between EtOAc and H₂O and the organic subjected to the standard work up procedure to yield the crude product, which was then purified by flash chromatography (C18, H₂O→MeOH gradient) to yield the pentyl ether (8) (40.3 mg, 0.11 mmol, 52% over 2 steps).

8: immobile oil; [α]²⁵_D +71.6 (c 0.787, CHCl₃); IR ν_max 3523, 2924, 2857, 1456, 1056 cm⁻¹; ¹H NMR (CDCl₃, 600 MHz) δ 5.70 (s, 1H), 4.97 (d, 1H, J = 9.2 Hz), 3.84 (m, 1H), 3.75 (m, 1H), 3.67 (m, 1H), 3.17 (m, 1H), 2.17 (m, 1H), 2.07 (s, 3H), 2.06–2.02 (m, 2H), 1.97 (m, 1H), 1.80 (m, 2H), 1.73 (s, 3H), 1.68 (s, 3H), 1.46 (m, 2H), 1.40 (m, 2H), 1.34–1.31 (m, 1H), 1.30 (d, 3H, J = 6.8 Hz), 1.23–1.21 (m, 2H), 1.03 (d, 3H, J = 6.0 Hz), 0.95 (m, 1H), 0.95 (t, 3H, 7.2 Hz); ¹³C NMR (CDCl₃, 150 MHz) δ 144.8, 142.8, 135.6, 131.3, 129.8, 128.3, 126.7, 125.5, 73.6, 44.7, 40.2, 37.0, 34.0, 32.1, 30.1, 28.8, 28.2, 27.8, 25.4, 23.0, 22.6, 20.0, 17.5, 14.0, 12.7; APCIMS m/z 371 [M + H]⁺; HRMS-ES m/z [M + H]⁺ 371.2942 (calcd for C₂₅H₃₉O₂, 371.2945).

A stirred solution of 9 [19] (25.5 mg, 0.083 mmol) in dry THF (5 mL), under N₂, was allowed to react with an excess of NaH for 2 h. Iodomethane (200 μL, excess) was then added and the solution was left stirring at room temperature overnight. The reaction was then carefully quenched with MeOH (1 mL), and excess iodomethane was removed under a stream of N₂. The products were then partitioned between EtOAc and H₂O. The EtOAc fraction was subjected to the standard work up procedure to give the crude product which was purified by flash chromatography (silica, hexane→MTBE gradient) to yield the desired dimethyl ether (10) (18.2 mg, 0.055 mmol, 67%).

10: amorphous solid; [α]²⁵_D +69.5 (c 0.573, CHCl₃); IR ν_max 2924, 2861, 1460, 1069 cm⁻¹; ¹H NMR (CDCl₃, 600 MHz) δ 4.99 (d, 1H, J = 9.2 Hz), 3.88 (s, 3H), 3.81 (s, 3H), 3.73 (q, 1H, J = 8.8 Hz), 3.26 (m, 1H), 2.15–2.05 (m, 3H), 2.11 (s, 3H), 2.00 (m, 1H), 1.76 (br s, 3H), 1.71 (br s, 3H), 1.39 (m, 1H), 1.31–1.23 (m, 2H), 1.28 (d, 3H, J = 6.9 Hz), 1.07 (d, 3H, J = 6.0 Hz), 0.98 (m, 1H); ¹³C NMR (CDCl₃, 150 MHz) δ 149.4, 149.0, 135.3, 134.0, 133.0, 130.9, 128.5, 128.4, 60.1, 59.8, 44.0, 40.1, 37.3, 34.0, 31.3, 28.3, 27.5, 25.4, 24.3, 20.0, 17.5, 12.1; APCIMS m/z 329 [M + H]⁺; HRMS-ES m/z [M + H]⁺ 329.2459 (calcd for C₂₂H₃₃O₂, 329.2475).

Triflic anhydride (6 mmol) in DCM was added to a stirred, ice-cooled solution of 9 (178 mg, 0.56 mmol) and Hunig’s base (2 mL) in dry toluene (8 mL). The reaction was stirred under N₂, and allowed to warm to room overnight before being partitioned between DCM and aqueous HCl (1 N). The organic phase was subjected to the standard work up procedure to yield the crude triflate; purification was achieved by flash chromatography (silica, hexane→EtOAc gradient) to yield the triflate (11, 177 mg, 0.4 mmol, 70%, ~90% pure by ELSD LCMS). A portion of the product was further purified by RP-HPLC (Phenomenex, phenylhexyl, 5 μm, 250 × 10 mm, 4.0 mL/min) using a gradient of MeOH/H₂O (9:1 for 1 min, then to 10:0 over 1–4 min; eluted across 9.9 to 10.25 min).

11: amorphous solid; [α]²⁵_D +62.8 (c 0.205, CHCl₃); IR ν_max 2928, 2869, 1416, 1206, 1139 cm⁻¹; ¹H NMR (CDCl₃, 600 MHz) δ 4.93 (d, 1H, J = 9.3 Hz), 3.74 (s, 3H), 3.72 (m, 1H,), 3.23 (dd, 1H, J = 7.2, 14.6 Hz), 2.18–2.06 (m, 3H), 2.12 (s, 3H), 2.01 (m, 1H), 1.74 (br s, 3H), 1.70 (br s, 3H), 1.37 (m, 1H), 1.26–1.21 (m, 2H), 1.25 (d, 3H, J = 6.9 Hz), 1.04 (d, 3H, J = 5.9 Hz), 0.97 (m, 1H); ¹³C NMR (CDCl₃, 150 MHz) δ 147.9, 140.1, 139.7, 136.7, 132.6, 129.7 (2 × C), 129.5, 118.7 (q, J = 320 Hz), 117.6, 60.8, 44.0, 39.8, 37.4, 33.9, 30.9, 28.6, 27.0, 25.4, 23.3, 19.8, 17.6, 12.5; APCIMS m/z 447 [M + H]⁺; HRMS-ES m/z [M + H]⁺ 447.1804 (calcd for C₂₂H₃₀F₃O₄S, 447.1811).

A stirred solution of 9 (25.5 mg, 0.083 mmol) in dry THF (5 mL), under N₂, was allowed to react with an excess of NaH for 2 h. Dimethylcarbamoyl chloride (200 μL, excess) was then added and the solution was stirred overnight. The reaction was then carefully quenched with MeOH (1 mL) and the products were portioned between EtOAc and H₂O. The EtOAc fraction was subjected to the standard work up procedure to give the crude product which was purified by flash chromatography (silica, hexane→MTBE gradient) to yield the carbamate (12, 13.7 mg, 0.035 mmol, 43%).

Carbamate (12): amorphous solid; [α]²⁵_D +110.2 (c 0.07, CHCl₃); IR ν_max 2924, 2861, 1723 (CO), 1451, 1386, 1164 cm⁻¹; ¹H NMR (CDCl₃, 600 MHz) δ 4.97 (d, 1H, J = 9.2 Hz), 3.70 (s, 3H), 3.71–3.67 (m, 1H), 3.16 (s, 3H), 3.04 (s, 3H), 3.03 (m, 1H), 2.14–2.10 (m, 2H), 2.07 (s, 3H), 2.07–2.02 (m, 2H), 1.98–1.94 (m, 1H) 1.73 (br s, 3H), 1.68 (br s, 3H), 1.34–1.16 (m, 6H), 1.01 (d, 3H, J = 6.1 Hz), 0.99 (m, 1H); ¹³C NMR (CDCl₃, 150 MHz) δ 154.6, 148.6, 140.9, 136.4, 135.3, 130.6, 129.6, 128.7, 128.3, 60.5, 44.2, 40.0, 37.3, 36.8, 36.4, 33.7, 31.5, 28.6, 27.3, 25.4, 23.8, 19.9, 17.5, 12.3; APCIMS m/z 386 [M + H]⁺; HRMS-ES m/z [M + Na]⁺ 408.2505 (calcd for C₂₄H₃₅NO₃Na, 408.2509).

A solution of 2-methyl-3-buten-2-ol (640 mg, 7.4 mmol) in DCM (3 mL) was added dropwise to a stirred mixture of 2,6-dimethoxyphenol (13, 612 mg, 3.97 mmol) and TsOH (19 mg, cat) in DCM/MeOH (3:1, 30 mL). After stirring for 96 h at room temperature the solution was refluxed for 20 h and then partitioned between H₂O and EtOAc. The organic phase was subjected to the standard work up procedure to yield a crude oil (793 mg) which was subjected to flash chromatography (C18, H₂O→MeOH gradient) to yield the mono-prenylated product (14, 239 mg, 1.08 mmol, 27%), the di-prenylated product (15, 96 mg, 0.33 mmol, 8%), and the triprenylated product (16, 20 mg, 0.055 mmol, 1.4%) along with recovered starting material (321 mg, 52%).

2,6-Dimethoxy-3-(3-methylbut-2-enyl)phenol (14): oil; IR ν_max 3456, 2931, 2835, 1493, 1288, 1090 cm⁻¹; ¹H NMR (CDCl₃, 600 MHz) δ 6.63 (d, 1H, J = 8.5 Hz), 6.60 (d, 1H, J = 8.4 Hz), 5.53 (s, 1H, OH), 5.27 (m, 1H, H-8), 3.863 (s, 3H, OMe), 3.861 (s, 3H, OMe), 3.30 (d, 2H, J = 7.3 Hz); 1.73 (s, 6H, H-10 & H-11); ¹³C NMR (CDCl₃, 150 MHz) δ 145.9, 145.2, 138.5, 132.0, 127.8, 123.1, 119.1, 106.4, 60.4, 56.2, 28.0, 25.7, 17.7; APCIMS m/z 223 [M + H]⁺; HRMS-ES m/z [M + Na]⁺ 245.1142 (calcd for C₁₃H₁₈O₃Na, 245.1148).

2,6-Dimethoxy-3,4-bis(3-methylbut-2-enyl)phenol (15): pale yellow oil; IR ν_max 3440, 2964, 2912, 1854, 1497, 1309, 1116 cm⁻¹; ¹H NMR (CDCl₃, 600 MHz) δ 6.50 (s, 1H, H-5), 5.23 (m, 1H, olefinic), 5.07 (m, 1H, olefinic), 3.85 (s, 3H, OMe), 3.83 (s, 3H, OMe), 3.32 (d, 2H, J = 6.5 Hz), 3.25 (d, 2H, J = 6.5 Hz), 1.77 (s, 3H), 1.75 (s, 3H), 1.71 (s, 3H), 1.68 (s, 3H). ¹³C NMR (CDCl₃, 150 MHz) δ 145.5, 145.4, 136.7, 132.1, 131.0 (2 × C), 126.0, 123.7, 123.4, 107.6, 60.6, 56.1, 31.4, 25.7, 25.6, 25.1, 17.9, 17.8; APCIMS m/z 291 [M + H]⁺; HRMS-ES m/z [M + Na]⁺ 313.1763 (calcd for C₁₈H₂₆O₃Na, 313.1774).

2,6-Dimethoxy-3,4,5-tris(3-methylbut-2-enyl)phenol (16): colorless oil; IR ν_max 3400, 2964, 2912, 2855, 1456, 1087 cm⁻¹; ¹H NMR (CDCl₃, 600 MHz) δ 5.49 (s, 1H, OH), 5.09 (m, 2H, olefinic), 4.97 (m, 1H, olefinic), 3.80 (s, 6H, OMe), 3.32 (d, 4H, J = 6.5 Hz), 3.23 (d, 2H, J = 6.0 Hz), 1.74 (s, 6H), 1.68 (broad overlapping singlets, 12H). ¹³C NMR (CDCl₃, 150 MHz) δ 144.2, 140.1, 131.2, 131.0, 130.7, 129.5, 123.9, 123.8, 60.9, 27.8 (2 × C), 25.6, 25.55, 25.53, 17.90, 17.91; APCIMS m/z 359 [M + H]⁺; HRMS-ES m/z [M + Na]⁺ 381.2409 (calcd for C₂₃H₃₄O₃Na, 381.2400).

The glycosyl donor was synthesized from 2,3,4,6-tetra-O-benzoyl-α-D-galactopyranosyl bromide, which was freshly prepared using a previously described method [28]. The glycosyl bromide (2.05 g, 3.11 mmol) was hydrolyzed and subsequently reacted with trichloroacetonitrile according to existing methodology to provide the glycosyl donor 21 (704 mg, 0.951 mmol, 31% over two steps) [29].

2,6-Dimethoxyphenol (13, 30.0 mg, 0.195 mmol) and freshly prepared benzoylated glycosyl donor (21, 151.0 mg, 0.229 mmol) were dissolved in 8 mL anhydrous DCM and stirred with 3Å molecular sieves (500 mg) under a N₂ atmosphere for 20 min. Afterwards, the mixture was cooled to −78 °C and BF₃·Et₂O (0.206 mmol) was added. After stirring for 2.5 h at −78 °C, the glycosyl donor was completely consumed as indicated by TLC. The reaction was quenched with Et₃N (100 μL, excess), filtered, diluted with EtOAc, and partitioned with H₂O. The EtOAc phase was recovered and concentrated in vacuo to provide the crude galactoside. Purification by flash chromatography (silica, hexane→MTBE gradient), yielded 17 (101.9 mg, 0.139 mmoles, 71%).

17: white solid; [α]²⁵_D +67.4 (c 0.1917, CH₂Cl₂); IR ν_max 3065, 2962, 2939, 2838, 1726, 1601, 1479, 1258, 1109, 1069, 709 cm⁻¹; ¹H NMR (C₆D₆, 600 MHz) δ 8.11 (d, 2H, J = 7.8 Hz), 8.10 (d, 2H, J = 7.8 Hz), 8.07 (d, 2H, J = 7.8 Hz), 8.00 (d, 2H, J = 7.8 Hz), 7.10 (t, 1H, J = 7.5 Hz), 7.06 (t, 1H, J = 7.5 Hz), 7.03 (t, 2H, J = 7.7 Hz), 7.01 (t, 1H, J = 7.5 Hz), 6.92 (t, 2H, J = 7.6 Hz), 6.91 (t, 2H, J = 7.6 Hz), 6.85 (t, 1H, J = 7.5 Hz), 6.78 (t, 1H, J = 8.4 Hz, H-4), 6.73 (t, 2H, J = 7.7 Hz), 6.60 (dd, 1H, J = 10.4, 7.9 Hz, H-2′), 6.23 (d, 2H, J = 8.4 Hz, H-3), 6.14 (dd, 1H, J = 3.6, 1.2 Hz, H-4′), 5.80 (dd, 1H, J = 10.4, 3.6 Hz, H-3′), 5.44 (d, 1H, J = 7.9 Hz, H-1′), 4.67 (dd, 1H, J = 11.3, 6.6 Hz, H-6′), 4.36 (dd, 1H, J = 11.3, 6.6 Hz, H-6′), 3.66 (app. ddd, 1H, J = 6.6, 1.2 Hz, H-5′), 3.25 (s, 6H, OMe); ¹³C NMR (C₆D₆, 150 MHz) δ 165.9, 165.8, 165.7, 165.4, 153.9 (C-2), 136.1 (C-1), 133.2, 133.1, 133.0, 132.7, 130.8, 130.5, 130.3, 130.1, 130.1, 130.0, 129.7, 129.6, 128.8, 128.6, 128.4, 128.3, 124.6 (C-4), 106.1 (C-3), 103.1 (C-1′), 72.7 (C-3′), 71.9 (C-5′), 71.4 (C-2′), 68.8 (C-4′), 62.3 (C-6′), 55.9 (OMe); HRMS-ES m/z [M + Na]⁺ 755.2088 (calcd for C₄₂H₃₆O₁₂Na, 755.2099).

Compound 17 (40.3 mg, 55.0 μmol) was stirred with K₂CO₃ (38.8 mg) in 3 mL MeOH:MTBE (5:1) for 20 h. After deprotection was completed, as indicated by TLC (silica, hexane/MTBE), the reaction mixture was diluted with H₂O and desalted by solid phase extraction (C18, 1:19 MeOH:H₂O). The product was eluted with MeOH and purified by flash chromatography (C18, H₂O→MeOH gradient) to provide 18 (12.5 mg, 39.5 μmol, 72%).

18: colorless solid; [α]²⁵_D −15.9 (c 0.8333, CH₂Cl₂); IR ν_max 3388, 3008, 2940, 2841, 1599, 1480, 1256, 1108, 1072 cm⁻¹; ¹H NMR ((CD₃)₂SO, 600 MHz) δ 6.97 (t, 1H, J = 8.4 Hz, H-4), 6.66 (d, 2H, J = 8.4 Hz, H-3), 4.83 (d, 1H, J = 7.6 Hz, H-1′), 4.79–4.74 (m, 2H, C-2′/3′-OH), 4.50-4.44 (m, 2H, C-4′/6′-OH), 3.74 (s, 6H, OMe), 3.67 (app. t, 1H, J = 3.3 Hz, H-4′), 3.56–3.52 (m, 1H, H-2′), 3.56–3.52 (m, 1H, H-6′), 3.37–3.33 (m, 1H, H-3′), 3.36–3.32 (m, 1H, H-6′), 3.27 (app. t, 1H, J = 6.3 Hz, H-5′); ¹³C NMR ((CD₃)₂SO, 150 MHz) δ 153.0 (C-2), 134.9 (C-1), 123.7 (C-4), 106.6 (C-3), 103.5 (C-1′), 75.5 (C-5′), 73.2 (C-3′), 71.4 (C-2′), 67.9 (C-4′), 60.1 (C-6′), 56.4 (OMe); HRMS-ES m/z [M + Na]⁺ 339.1043 (calcd for C₁₄H₂₀O₈Na, 339.1050).

The glycosylation procedure, described for the synthesis of 17, was repeated using the benzoylated glycosyl donor (21, 175.6 mg, 0.266 mmol) and 14 (44.6 mg, 0.201 mmol) as the glycosyl accepting substrate. The reaction crude was separated by flash chromatography (silica, hexane→MTBE gradient) to provide 19 (149.1 mg, 0.186 mmol, 93%).

19: pale yellow solid; [α]²⁵_D +75.9 (c 0.225, CH₂Cl₂); IR ν_max 3063, 2966, 2936, 1726, 1602, 1493, 1451, 1094, 1069, 709 cm⁻¹; ¹H NMR (CD₃OD, 600 MHz) δ 8.13 (d, 2H, J = 7.7 Hz), 7.90 (d, 4H, J = 7.8 Hz), 7.75 (d, 2H, J = 7.8 Hz), 7.67 (t, 1H, J = 7.7 Hz), 7.57 (t, 1H, J = 7.7 Hz), 7.54 (t, 2H, J = 7.7 Hz), 7.51 (t, 1H, J = 7.7 Hz), 7.45 (t, 1H, J = 7.6 Hz), 7.40 (t, 2H, J = 7.7 Hz), 7.36 (t, 2H, J = 7.7 Hz), 7.25 (t, 2H, J = 7.7 Hz), 6.80 (d, 1H, J = 8.6 Hz, H-4), 6.54 (d, 1H, J = 8.6 Hz, H-5), 6.03 (dd, 1H, J = 10.3, 8.0 Hz, H-2′), 6.00 (app. d, 1H, J = 3.4 Hz, H-4′), 5.78 (dd, 1H, J = 10.3, 3.4 Hz, H-3′), 5.64 (d, 1H, J = 8.0 Hz, H-1′), 5.14 (app. t, 1H, J = 7.4 Hz, H-8), 4.58 (dd, 1H, J = 10.8, 7.0 Hz, H-6′), 4.54 (app. dd, 1H, J = 7.0, 4.9 Hz, H-5′), 4.47 (dd, 1H, J = 10.8, 4.9 Hz, H-6′), 3.75 (s, 3H, C2-OMe), 3.53 (s, 3H, C6-OMe), 3.17-3.09 (m, 2H, H-7), 1.67 (s, 3H, H-10), 1.64 (s, 3H, H-11); ¹³C NMR (CD₃OD, 150 MHz) δ 167.5, 167.3, 167.1, 166.8, 152.7 (C-2), 152.6 (C-6), 139.9 (C-1), 134.9, 134.6, 134.5, 134.4, 134.3, 132.9, 131.0, 131.0, 130.9, 130.7, 130.7, 130.6, 130.6, 130.5, 130.2 (C-9), 129.9, 129.6, 129.6, 129.5, 129.5, 129.1 (C-3), 129.0, 128.8, 128.7, 125.9 (C-4), 124.4 (C-8), 109.0 (C-5), 103.2 (C-1′), 73.6 (C-3′), 73.0 (C-5′), 72.3 (C-2′), 70.2 (C-4′), 63.7 (C-6′), 61.7 (2-OMe), 56.6 (6-OMe), 29.0 (C-7), 25.9 (C-10), 17.9 (C-11); HRMS-ES m/z [M + Na]⁺ 823.2754 (calcd for C₄₇H₄₄O₁₂Na, 823.2725).

The deprotection procedure, described for the synthesis of 18, was repeated with 19 (38.5 mg, 48.1 μmol) and K₂CO₃ (33.4 mg). The deprotected glycoside was purified with the method described for 17 to yield 20 (13.9 mg, 36.2 μmol, 75%).

20: colorless solid; [α]²⁵_D −6.44 (c 0.6917, CH₃OH); IR ν_max 3392, 2967, 2914, 1602, 1494, 1463, 1442, 1090 cm⁻¹; ¹H NMR ((CD₃)₂SO, 600 MHz) δ 6.79 (d, 1H, J = 8.5 Hz, H-4), 6.70 (d, 1H, J = 8.5 Hz, H-5), 5.20 (app. t, 1H, J = 7.3 Hz, H-8), 4.96 (d, 1H, J = 4.5 Hz, C-2′-OH), 4.92 (d, 1H, J = 7.6 Hz, H-1′), 4.80 (d, 1H, J = 5.3 Hz, C-3′-OH), 4.49 (m, 1H, C-4′-OH), 4.47 (m, 1H, C-6′-OH), 3.76 (s, 3H, C6-OMe), 3.72 (s, 3H, C2-OMe), 3.68 (app. dd, 1H, J = 3.5 Hz, H-4′), 3.59–3.55 (m, 1H, H-2′), 3.55–3.52 (m, 1H, H-6′), 3.39–3.36 (m, 1H, H-3′), 3.35–3.32 (m, 2H, H-6′), 3.29 (app. t, 1H, J = 6.2 Hz, H-5′), 3.24–3.15 (m, 2H, H-7), 1.68 (s, 3H, H-10), 1.67 (s, 3H, H-11); ¹³C NMR ((CD₃)₂SO, 150 MHz) δ 151.6 (C-2), 151.1 (C-6), 138.8 (C-1), 131.3 (C-9), 127.4 (C-3), 123.6 (C-4), 123.4 (C-8), 108.9 (C-5), 103.2 (C-1′), 75.5 (C-5′), 73.2 (C-3′), 71.4 (C-2′), 67.9 (C-4′), 60.7 (6-OMe), 60.1 (C-6′), 56.5 (2-OMe), 27.7 (C-7), 25.5 (C-10), 17.6 (C-11); HRMS-ES m/z [M + Na]⁺ 407.1687 (calcd for C₁₉H₂₈O₈Na, 407.1676).