Synthetic studies towards Zetekitoxin AB: preparation of 4,5-epi-11-hydroxy-saxitoxinol

Abstract

A concise synthesis of 4,5-epi-11-hydroxy-saxitoxinol utilizing D-ribose to direct an asymmetric Mannich reaction. This approach allows many modes of reactivity, which can be used to access various analogs of saxitoxin.

1. Introduction

Saxitoxin (2) and its analogs have attracted the interest of organic chemists due to its dense, heteroatom-rich structure.^1–10 Zetekitoxin AB (1), isolated from Panamanian golden frog (Atelopus zeteki), was found to be an analog of saxitoxin however, significantly more biologically potent.¹¹ Due to the endangerment of the Panamanian golden frog there is no source of Zetekitoxin AB, preventing further studies. Due to the complexity, potency and rarity of this molecule we were attracted to tackling the challenge of its synthesis.

Our retrosynthetic analysis reveals that Zetekitoxin AB could arise from core structure 4. This substrate could arise from the condensation of two guanidines on a central carbonyl and the nucleophilic ring-closure on carbon-10, displacing a leaving group (Scheme 1). Ketone 5 could be derived from protected diamino ester 6, which in turn could be formed from a Mannich-type reaction on ribose-derived imine 7. The Mannich precursor could be formed from D-ribose (8) in a few steps.

Scheme 1.

Retrosythetic analysis of zetekitoxin AB

2. Results and Discussion

D-Ribose was treated with sulfuric acid, benzyl alcohol and acetone to give the corresponding O-benzyl-protected ribose acetal, which was oxidized with TEMPO and TCC to give aldehyde 9 (Scheme 2). This aldehyde was converted to several benzyl imines or oximes and then added to a solution of the enolate of glycine template 11, resulting only in the recovery of starting material.

Scheme 2.

Preparation of Mannich product 12 from D-Ribose (8)

These results indicated that imines or oximes were not electrophilic enough, so more reactive imines were sought out. Since carbamates are good electron-withdrawing groups and versatile amine protecting groups it was thought that N-acylimines would constitute a suitable Mannich substrate. A literature survey revealed that the N-acylimines are generally not stable, but they could be masked and generated in situ. Amido sulfones, in particular, seemed to be the most promising because they are stable and can be prepared with mild conditions.¹²

When ribose-derived aldehyde 9 was treated with tert-butyl carbamate, phenylsulfinic acid, and anhydrous magnesium sulfate in dichloromethane, amido sulfone 10 was generated in good yield (Scheme 2).¹³ Amido sulfone 10 was added to the enolate of glycine template 11 giving desired Mannich product 12 in good yield and as a single diastereomer. The absolute stereochemistry of product 12 was determined by X-ray crystallography.

With Mannich product 12 in hand, the Boc group was removed via Ohfune’s conditions followed by the hydrolysis of the benzophenone imine to give the diamine. This diamine was then guanidinylated with reagent 13, mercuric chloride and triethylamine in DMF to give bisguanidine 14 (Scheme 3). Hydrogenation of compound 14, followed by reduction of the lactol gave the diol, which was selectively protected with TBSOTf. The resulting alcohol was then oxidized to the ketone using Dess-Martin’s periodinane giving ketone 15, which embodies all the functionality to make the core of Zetekitoxin AB.

Scheme 3.

Elaboration of Mannich product 12 to ketone 15

With ketone 15 in hand the next step was removal of the methyl carbamate protecting groups, followed by treatment with an acid to form bicyclic species 16 (Scheme 4). This bicyclic substance could then be submitted to Mitsunobu conditions giving compound 17. This promising synthetic approach was abandoned at this stage because another approach, being run in parallel, showed greater potential.

Scheme 4.

Hypothetical formation of tricycle

Starting with Mannich product 12, the benzophenone imine was hydrolyzed and then treated with methyl chloroformate to give methyl carbamate 18 (Scheme 5). The ethyl ester was then reduced with lithium borohydride giving a mixture of oxazolidinone 19 and the primary alcohol. This mixture could be converted to oxazolidinone 19 by heating in a solution of potassium hydroxide in methanol.

Scheme 5.

Synthesis of oxazolidinone 19

Exposure of oxazolidinone 19 to Ohfune’s conditions afforded the corresponding amine, which was then guanidinylated with reagent 20 giving oxazolidinone 21 in good yield (Scheme 6). The benzyl ether was then removed via hydrogenation at high pressure with Pearlman’s catalyst to give lactol 22.

Scheme 6.

Preparation of lactol 22

Installation of an amino group at the lactol position was not as straightforward as expected. Initial attempts for a one pot reductive amination failed, so a two-step procedure was utilized. Treatment of the lactol 22 with hydroxylamine-HCl and sodium acetate afforded the corresponding oxime. The oxime proved recalcitrant to reduction with a number of hydride reducing reagents or by hydrogenation with either palladium or platinum catalysts. The best conditions found to reduce the oxime was hydrogenation over Raney nickel (Scheme 7). Unfortunately, the yield of amine 23 was only 25% over the two steps and suffered from poor reproducibility.

Scheme 7.

Two step installation of amine.

After screening many conditions, a one-pot reductive amination protocol was developed. When lactol 22 was stirred with a benzyl amine and sodium cyanoborohydride in methanol there was no reaction. It was discovered that adding ~150 equivalents of acetic acid to the reaction gave the desired amine, in a reasonable amount of time (Scheme 8). Many different benzylamines were tested however the o-nitrobenzyl (oNB) was the only protecting group that proved later, could be removed efficiently.

Scheme 8.

Reductive amination with oNB-amine

Amines 23 and 24 were elaborated to nine-membered ring species 26 and 27 in a one-pot reaction (Scheme 9). This was accomplished by the treatment of the amine with reagent 25 followed by a mercury-promoted cyclization and ring opening of the resulting oxazolidinone with ammonia in water. These nine-membered ring structures are reminiscient of the intermediate in DuBois’s elegant synthesis of saxitoxin.^5,6 Alcohols 26 and 27 were efficiently oxidized to ketone 28 and 29 with Dess-Martin periodinane. Ketone 28 spontaneously undergoes ring contraction to give hemiaminal 30, as a mixture of diastereomers, due to the lack of the benzyl-protecting group. Irradiating compound 29 with UV light cleanly afforded the desired free guanidine which cyclized to the hemiaminal 30.

Scheme 9.

Installation and cyclization of the isothiourea

The first reaction conditions tested to enact the global deprotection and cyclization to the tricycle was 0.5M B(TFA)₃ in TFA, which had literature precedence.⁶ Unfortunately, these conditions failed to give the desired result. To simplify this complex reaction a stepwise approach was taken allowing each individual product to be characterized. Treatment of the mixture of 29/30 with 6M HCl in methanol gave a single product. This product showed loss of the Boc groups, the acetonide and a water molecule by NMR and MS analysis. The NMR spectrum indicated the presence of an olefin, but no hemiaminal or aminal. Since the identity of the compound couldn’t be determined easily, it was treated with TFA. This product still contained the olefin and showed the loss of another water molecule. Since hemiaminal 30 is presumably in equilibrium with ketone 29, two different pathways can be imagined (Scheme 10). Hemiaminal 30 could lose both Boc groups and the acetonide followed by dehydration to form enamine 31. A very similar pathway could occur with ketone 29 giving enamine 35. These enamines when treated with TFA also could undergo very similar reactions. Elimination of water could give reactive intermediates 32 or 36, which could then cyclize to compound 33 or 37. These compounds are structurally similar and distinguishing between them by NMR proved to be not straightforward.

Scheme 10.

Potential cyclization pathways

To identify the pathway that is operative we endeavored to narrow down the possibilities (Scheme 11). Leaving the o-NB protecting group intact until after the cyclizations eliminated the possibility of pathway A. Treatment of compound 28 with 6M HCl/MeOH followed by uv irradiation gave compound 35 which matched the previously made substance. Furthermore, treatment of compound 35 with TFA led to compound 37. These results indicated that pathway B was the preferred route.

Scheme 11.

Determination of the cyclization pathway

This result although interesting was not synthetically useful for this synthesis. To avoid this side-reaction we speculated that if we could prevent z the hemiaminal 30 from opening to the 9-member ring we could force the compound through an alternative pathway, which might lead to the desired product. We envisioned that using an oxophilic reagent the hemiaminal could be stabilized. We decided to reexamine the B(TFA)₃ in TFA conditions because boron is very oxophilic and should be capable of trapping hemiaminal 30. It was found that a slight modification in the procedure gave the desired result. Originally, the compound was cooled, dissolved in TFA and the B(TFA)₃ was added. It is presumed that the Boc deprotection is so fast that the more stable hemiaminal 34 was probably forming before the boron reagent was added. Instead, the B(TFA)₃ in TFA was added directly to a cooled mixture of 29/30, which after stirring for 15h. resulted in the formation of tricycle 40 and enamine 41.

This result was exciting, but the stereogenic centers at C-8 and C-9 in compound 40 possess the wrong configuration. It was reasoned that the enamine (41) formation might be reversible. If this is the case it could lead to epimerization of the incorrect stereogenic center and give the desired tricycle 43 which should be more stable than tricycle 40, since its steric interactions between the newly formed 5-member spiro-guanidine ring and the proximal substituents could be obviated (Scheme 13).

Scheme 13.

Plan to epimerize C-5

Mixture 29/30 was allowed to stir for 48h, but instead of giving the desired tricycle, compound 44 was formed exclusively. Unfortunately, this reaction proved to be irreversible.

The undesirable elimination/cyclization steps could theoretically be slowed down with cooling, but unfortunately when the same reaction was stirred for 48h. at 4°C the result was a nearly 1:1 mixture of tricycle 40 and compound 44. This result indicated that tricycle 40 was fairly stable and only a small fraction converted to enamine 41, which easily eliminates water irreversibly giving the undesirable compound 44. Compound 40 is of 4,5-epi-11-hydroxy-saxitoxinol.

3. Conclusions

In conclusion, we have explored a strategy for the synthesis of Saxitoxin analogs utilizing a diastereoselective Mannich-type reaction on a ribose-derived imine. This approach allows for the efficient assembly of the Saxitoxin backbone in four steps, which can be elaborated to the tricyclic core of Saxitoxin in a total of fifteen steps. Future efforts will focus on rectifying stereochemical issues at C-8 in the early stages of the synthesis.

4. Experimental

4.1 General

Unless otherwise noted, materials were obtained from commercial sources and used without purification. All reactions requiring anhydrous conditions were performed under positive pressure of argon using flame dried glassware. Dichloromethane, diisopropylamine, triethylamine, tetrahydrofuran, dimethylformamide, diethyl ether, and toluene were degassed with argon and passed through a solvent purification system (J.C. Meyer of Glass Contour) containing either alumina or molecular sieves. Flash chromatography was performed on Merck silica gel Kieselgel 60 (230–400 mesh) from EM science with the indicated solvent.

¹H-NMR spectra were recorded on Varian, 300, 400, or 500 MHz spectrometers as indicated. The chemical shifts (δ) of proton resonances are reported relative to CDCl3, DMSO-d5, D₂O, CD₃OD, and acetone-d6 using the following format: chemical shift [multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet), coupling constant(s) (J in Hz), integral]. ¹³C-NMR spectra were recorded at 100 or 125 MHz. The chemical shifts of carbon resonances are reported relative to the deuterated solvent peak, except D₂O, which was internally referenced.

Infrared spectra were recorded on a Bruker Tensor 27 FTIR spectrometer. All absorptions are reported in cm⁻¹. Spectra were recorded as films deposited from deuterated NMR solvent solutions on NaCl plates followed by solvent evaporation. Peaks reported in the IR are described using the following conventions: w = weak, m = medium, s = strong, vs = very strong, sh = shoulder, and br = broad.

Mass spectra were obtained at the Colorado State University CIF on a Fisons VG Autospec. Optical rotations were obtained with a 1mL, 1 dm cell on a Rudolf Research Autopol III polarimeter operating at 589 nm in CHCl₃. HPLC separations were obtained on a Waters 600E HPLC system using the indicated column and eluent conditions.

4.2 Experimental procedures

((3aR,4R,6R,6aR)-6-(benzyloxy)-2,2-dimethyltetrahydrofuro[ 3,4-d][1,3]dioxol-4-yl)methanol (8a)

To a suspension of D-ribose (150 g) in benzyl alcohol (990 mL) and acetone (600 mL) was added sulfuric acid (12 mL) drop wise. The mixture was heated to 70–75°C and was allowed to stir for 4–5h. The reaction was allowed to cool to room temperature and was neutralized with Et₃N. The acetone was removed under vacuum and the solution was washed with H₂O (3x). The benzyl alcohol was removed under high vacuum. The product was crashed out with Et₂O and the solids were washed with 20% EtOAc/hexanes to give 141.7 grams (51%) of an off white solid. ${[\alpha ]}_D^{25}=-96$ (c 0.44, CHCl₃) ¹H-NMR (400 MHz; CDCl₃): δ 7.38-7.30 (m, 5H), 5.18 (s, 1H), 4.86 (d, J = 5.9 Hz, 1H), 4.77 (d, J = 11.6 Hz, 1H), 4.67 (d, J = 5.9 Hz, 1H), 4.58 (d, J = 11.6 Hz, 1H), 4.45 (t, J = 2.9 Hz, 1H), 3.72 (dd, J = 12.5, 2.4 Hz, 1H), 3.63 (dd, J = 12.5, 3.6 Hz, 1H), 3.11 (s, 1H), 1.48 (s, 3H), 1.32 (s, 3H). 13-C NMR (101 MHz; cdcl3): δ 136.5, 128.8, 128.41, 128.35, 112.3, 108.2, 88.6, 86.1, 81.7, 70.3, 64.2, 26.5, 24.8 IR (Dep. CDCl₃): 3487 (br), 2930.17 (s), 2888 (s), 1454 (s), 1403 (s), 1375 (s).

tert-butyl ((R)-((3aS,4S,6R,6aR)-6-(benzyloxy)-2,2-dimethylte trahydrofuro[3,4-d][1,3]dioxol-4-yl)(phenylsulfonyl)methyl)c arbamate (9)

To a solution of aldehyde 8a (36.9 g, 1 eq.) in CH₂Cl₂ (1.3L) was added t-butyl carbamate (18.3 g, 1.2 eq.), phenylsulfinic acid (28.3 g, 1.5 eq.), and magnesium sulfate (19 g). This mixture was allowed to stir overnight and was filtered through glass wool. The solution was then evaporated in vacuo and the product was crashed out with ether (500mL) to give 42.4 g (62%) of 9 as a white powder and taken on crude.^{1 1}H-NMR (400 MHz; CDCl₃): δ 7.81 (s, J = 8.2, 1.4 Hz, 1H), 7.74 (d, J = 7.4 Hz, 1H), 7.47 (d, J = 5.1 Hz, 2H), 7.40-7.23 (m, 6H), 7.03 (dd, J = 7.5, 1.9 Hz, 1H), 4.99 (s, 1H), 4.88 (d, J = 5.9 Hz, 1H), 4.75 (d, J = 5.9 Hz, 1H), 4.53 (s, 1H), 4.44 (ddd, J = 11.6, 9.7, 1.9 Hz, 1H), 4.11 (dq, J = 10.6, 10.0, 7.0 Hz, 1H), 4.06 (d, J = 13.3 Hz, 1H), 3.90 (dq, J = 10.7, 7.1 Hz, 1H), 1.45 (s, 9H), 1.44 (s, 3H), 1.29 (s, 3H), 1.08 (t, J = 7.1 Hz, 3H). 13-C NMR (101 MHz; cdcl3): δ 170.2, 155.7, 137.0, 132.4, 130.0, 128.32, 128.23, 128.10, 127.9, 127.56, 127.49, 112.0, 108.0, 86.3, 85.4, 81.9, 79.5, 69.5, 64.8, 61.1, 54.3, 28.2, 26.2, 24.6, 13.9. IR (Dep. CDCl₃): 3428 (s), 2980(s), 2938(s), 1739(s), 1714(s), 1491(s),1714(s), 1491(s).

(2R,3R) -ethyl 3- ((3aR,4R,6R,6aR) -6- (benzyloxy) -2,2-dimethyltetrahydrofuro[3,4-d][1,3]dioxol-4-yl)-3-((tert-butoxycarbonyl)amino)-2-((diphenylmethylene) amino)propanoate (12)

To a solution of LDA (0.1M, 192.3 mL) in dry THF at −78°C was added a glycine 11 (5.14 g, 2 eq.) in THF (−78°C) dropwise. The resulting yellow solution was stirred for 1 hour and amido sulfone 9 (5 g, 1 eq.) in THF (−78°C) was added dropwise. The reaction was stirred 20 min. and then quenched with sat. NH₄Cl (100mL) and allowed to warm to rt. The organic phase was separated and washed with brine. The aqueous phase was extracted with EtOAc (3x, 50mL) followed by brine. The combined organic extracts were dried over sodium sulfate and evaporated. The resulting oil was dissolved in hexanes and was evaporated. This process was repeated until solids formed which were removed by filtration to give 12 (4.8 g, 78%) as a white powder, which could be recrystallized from hexanes. ${[\alpha ]}_D^{25}=-31$ (c 0.38, CHCl₃). ¹H-NMR (400 MHz; CDCl₃): δ7.74 (d, J = 7.4 Hz, 1H), 7.49-7.26 (m, 11H), 7.03 (dd, J = 7.5, 1.9 Hz, 2H), 6.15 (s, 1H), 5.80 (d, J = 9.7 Hz, 1H), 4.99 (s, 1H), 4.88 (d, J = 6.0 Hz, 1H), 4.75 (d, J = 5.9 Hz, 1H), 4.53 (s, 1H), 4.44 (ddd, J = 11.6, 9.7, 1.9 Hz, 1H), 4.10 (dq, J = 10.6, 7.1 Hz, 1H), 4.09 (d, J =12.0 Hz, 1H), 4.05 (d, J = 11.2 Hz, 1H), 4.03 (d, J = 11.6 Hz, 1H), 3.90 (dq, J = 10.7, 7.1 Hz, 1H), 1.45 (s, 9H), 1.44 (s, 3H), 1.29 (s, 3H), 1.08 (t, J = 7.1 Hz, 3H). 13-C NMR (101 MHz; cdcl3): δ 170.4, 155.9, 137.2, 132.5, 130.2, 128.60, 128.51, 128.42, 128.28, 128.07, 128.01, 127.81, 127.75, 127.67, 112.2, 108.2, 86.5, 85.5, 82.1, 79.7, 69.7, 65.0, 61.3, 54.5, 28.4, 26.4, 24.8, 14.1. IR (Dep. CDCl₃): 3428 (br), 2980 (vs), 2938 (vs), 1739 (vs), 1714 (vs), 1491 (s). HRMS (FAB⁺): Calcd for C₃₇H₄₅N₂O₈ [M+H]: 645.3176; Found 645.3172.

(2R,3R)-ethyl 2,3-diamino-3-((3aR,4R,6R,6aR)-6-(benzyloxy)-2,2-dimethyltetrahydrofuro[3,4-d][1,3]dioxol-4-yl)propanoate (12a)

Compound 12 (1000mg, 1eq.) was dissolved in CH₂Cl₂ (15.5mL) and 2,6-lutidine (0.67g, 4eq.) was added followed by TMSOTf (1.2g, 3.5eq.). The mixture was stirred 1 hr. and quenched with sat. NaHCO₃ (20mL). The mixture was extracted with EtOAc (3x, 50mL) dried over Na₂SO₄, evaporated in vacuo. The product was dissolved in THF (5mL) and 1M HCl (5mL) was added. The mixture was stirred for ~20 min. until complete by TLC. The mixture was washed with diethyl ether (3x, 20mL) to remove the benzophenone and then the aqueous layer was basified with NaHCO_3. The mixture was extracted with n-butanol (5x, 20mL) and the organic extracts were combined, dried over Na₂SO₄ and evaporated in vacuo to give diamine 12a (570mg, 97%), which was taken on crude. ¹H-NMR (300 MHz; CDCl₃): δ 7.34-7.29 (m, 5H), 5.19 (s, 1H), 4.97 (dd, J = 6.0, 1.1 Hz, 1H), 4.76 (d, J = 11.9 Hz, 1H), 4.71 (d, J = 6.0 Hz, 1H), 4.50 (d, J = 11.9 Hz, 1H), 4.14 (q, J = 7.1 Hz, 2H), 4.07 (dd, J = 10.7, 1.2 Hz, 1H), 3.98 (d, J = 2.0 Hz, 1H), 3.20 (dd, J = 10.7, 2.0 Hz, 1H), 1.49 (s, 3H), 1.33 (s, 3H), 1.17 (t, J = 7.1 Hz, 3H). HRMS (FAB⁺): Calcd. for C₁₉H₂₉N₂O₆ [M+H]: 381.2026; Found 381.2022.

(2R,3R)-ethyl 3-((3aR,4R,6R,6aR)-6-(benzyloxy)-2,2-dimethyl tetrahydrofuro[3,4-d][1,3]dioxol-4-yl)-2,3-bis(2-(tert-butoxyc arbonyl)-3-(methoxycarbonyl)guanidino)propanoate (14)

To a solution of diamine 12a (570mg, 1eq.) in DMF (15 ml) was added reagent 13 (1.12g, 1.1eq.), HgCl₂ (1.22g, 1.1eq.) and triethylamine (1.36g, 3.3 eq.). The mixture was stirred 30 min and the mercury salts were crashed out with ethyl acetate (25mL). The mixture was filtered and the filtrate was washed with H₂O. The organic layer was washed with brine (25mL), dried over Na₂SO₄ and solvents were removed in vacuo. The crude material was purified via flash chromatography to give product 14 (844mg, 72%). ${[\alpha ]}_D^{25}=-58$ (c 0.53, CHCl₃). ¹H-NMR (500 MHz; CDCl₃): δ 11.34 (s, 1H), 11.30 (s, 1H), 9.14 (d, J = 8.9 Hz, 1H), 8.85 (d, J = 9.7 Hz, 1H), 7.40 (d, J = 7.2 Hz, 2H), 7.33 (t, J = 7.3 Hz, 2H), 7.28 (d, J = 7.2 Hz, 1H), 5.40 (dd, J = 8.9, 2.1 Hz, 1H), 5.24 (s, 1H), 5.04 (td, J = 10.3, 1.6 Hz, 1H), 4.96 (d, J = 12.2 Hz, 1H), 4.77 (dd, J = 6.0, 1.0 Hz, 1H), 4.73 (d, J = 6.0 Hz, 1H), 4.59 (d, J = 12.2 Hz, 1H), 4.21 (dd, J = 10.7, 1.1 Hz, 1H), 4.14-4.04 (m, 2H), 3.76 (d, J = 5.7 Hz, 1H), 3.70 (s, 3H), 3.65 (s, 3H), 1.50 (s, 9H), 1.48 (s, 9H), 1.46 (s, 3H), 1.29 (s, 3H), 1.13 (t, J = 7.2 Hz, 3H). 13-C NMR (126 MHz; cdcl3): δ 170.1, 164.29, 164.21, 156.94, 156.88, 152.77, 152.67, 137.5, 128.5, 127.89, 127.73, 112.9, 108.1, 86.0, 85.4, 83.8, 83.6, 81.9, 70.1, 62.3, 53.86, 53.81, 52.9, 28.1, 26.6, 25.3, 13.9. IR (Dep. CDCl₃): 2987 (w), 2946 (w), 1722 (vs), 1644 (vs), 1620 (s), 1572 (s). HRMS (FAB⁺): Calcd for C₃₅H₅₃N₆O₁₄ [M+H]: 781.3620; Found 781.3625

(2R,3R)-ethyl 2,3-bis(2-(tert-butoxycarbonyl)-3-(methoxycarb onyl)guanidino)-3-((3aR,4R,6R,6aR)-6-hydroxy-2,2-dimethylt etrahydrofuro[3,4-d][1,3]dioxol-4-yl)propanoate (14a)

Benzyl ether 14 (540mg) was dissolved in 5:1 THF/MeOH (6mL) and 20% Pd(OH)₂/C (270mg) was added. Argon was bubbled through the suspension for 5 min. and the mixture was hydrogenated at 100 psi hydrogen for 24h. The mixture was then filtered through celite, evaporated, and purified via flash chromatography (30% EtOAc/Hexane) to give 404 mg (85%) of lactol 14a. ${[\alpha ]}_D^{25}=-1$ (c 0.54, CHCl₃) 1H NMR (500 MHz; CDCl3): δ 11.51 (s, 1H), 11.37 (s, 1H), 9.31 (d, J = 9.7 Hz, 1H), 8.97 (d, J = 9.0 Hz, 1H), 5.63 (s, 1H), 5.13 (dd, J = 8.9, 2.0 Hz, 1H), 4.93 (ddd, J = 11.6, 9.6, 2.3 Hz, 1H), 4.89 (d, J = 6.0 Hz, 1H), 4.68 (d, J = 5.9 Hz, 1H), 4.41 (s, 1H), 4.32 (q, J = 7.1 Hz, 2H), 4.12 (d, J = 11.4 Hz, 1H), 3.69 (s, 3H), 3.67 (s, 3H), 1.50 (s, 8H), 1.49 (s, 8H), 1.45 (s, 4H), 1.35 (t, J = 7.1 Hz, 3H), 1.31 (s, 3H). 13-C NMR (126 MHz; cdcl3): δ 169.3, 164.27, 164.11, 156.4, 155.5, 153.6, 152.5, 112.8, 104.0, 88.1, 85.6, 84.6, 83.7, 82.3, 62.6, 55.7, 52.91, 52.77, 52.3, 28.2, 26.8, 25.5, 14.1 IR (Dep. CDCl₃): 1724 (s), 1619 (m), 1440 (s), 1413 (s), 1370 (s). HRMS (FAB⁺): Calcd. for C₂₈H₄₇N₆O₁₄ [M+H]: 691.3150; Found 691.3147.

(2R,3R)-ethyl 2,3-bis(2-(tert-butoxycarbonyl)-3-(methoxycarb onyl)guanidino)-4-((4R,5S)-5-(((tert-butyldimethylsilyl)oxy)m ethyl)-2,2-dimethyl-1,3-dioxolan-4-yl)-4-hydroxybutanoate (14b)

To a solution of diol 14a (240mg, 1 eq.) in CH₂Cl₂ (20 mL) was added 2,6-lutidine (223mg, 6 eq.). The solution was cooled to −78°C and TBSOTf (446mg, 5 eq.) was added. The reaction was stirred 1 hr. and was quenched with sat. NH₄Cl. The mixture was extracted with ethyl acetate (3x, 10mL), dried over sodium sulfate and evaporated in vacuo. The crude compound was purified via column chromatography (10–30% EtOAc/Hexanes) to give TBS ether 14b (230 mg, 82%). ¹H-NMR (500 MHz; CDCl₃): δ 11.32 (s, 1H), 11.18 (s, 1H), 8.91 (d, J = 7.3 Hz, 1H), 8.80 (d, J = 6.1 Hz, 1H), 5.17 (t, J = 7.5 Hz, 1H), 4.82 (dd, J = 8.5, 7.0 Hz, 1H), 4.73 (q, J = 7.0 Hz, 1H), 4.64 (dd, J = 8.8, 6.3 Hz, 1H), 4.33-4.22 (m, 2H), 3.95 (dd, J = 10.7, 2.6 Hz, 1H), 3.85 (dd, J = 11.7, 5.0 Hz, 1H), 3.80-3.78 (m, 1H), 3.65 (s, 6H), 1.49 (s, 6H), 1.48 (s, 5H), 1.46 (t, J = 9.3 Hz, 3H), 1.37 (s, 3H), 1.32 (s, 3H), 0.89 (s, 9H), 0.08 (s, 3H), 0.07 (s, 3H). 13-C NMR (126 MHz; cdcl3): δ 171.6, 163.7, 163.5, 155.9, 155.6, 152.9, 152.7, 108.9, 83.98, 83.94, 78.1, 76.7, 61.5, 56.7, 54.6, 52.75, 52.66, 28.12, 28.06, 27.3, 26.00, 25.94, 25.2, 18.4, −5.2. HRMS (FAB⁺): Calcd. for C₃₄H₆₃N₆O₁₄Si [M+H]: 807.4172; Found 807.4168.

(2R,3R)-ethyl 2,3-bis(2-(tert-butoxycarbonyl)-3-(methoxycarb onyl)guanidino)-4-((4S,5S)-5-(((tert-butyldimethylsilyl)oxy)m ethyl)-2,2-dimethyl-1,3-dioxolan-4-yl)-4-oxobutanoate (15)

To a solution of compound 14b (49 mg, 1 eq.) in CH₂Cl₂ (5 mL) was added Dess-Martin periodinane (51.5 mg, 2 eq.). The solution turned cloudy over time and was allowed to stir for 2h. The reaction was quenched with 10% Na₂S₂O₃ (5mL) and was allowed to stir until the biphasic system turned clear. The mixture was extracted with EtOAc (3x, 10mL), dried over sodium sulfate and evaporated in vacuo to give ketone 15 (47 mg, 96%). ¹H-NMR (500 MHz; CDCl₃): δ 11.24 (s, 1H), 11.10 (s, 1H), 9.24 (d, J = 5.0 Hz, 1H), 9.00 (d, J = 6.5 Hz, 1H), 5.77 (d, J = 7.2 Hz, 1H), 5.72 (d, J = 8.0 Hz, 1H), 5.24 (d, J = 8.6 Hz, 1H), 4.46 (d, J = 8.6 Hz, 1H), 4.38-4.31 (m, 1H), 4.27-4.20 (m, 1H), 3.72 (dd, J = 11.7, 2.1 Hz, 1H), 3.67 (dd, J = 11.7, 2.1 Hz, 1H), 3.63 (s, 3H), 3.62 (s, 3H), 1.47 (s, 9H), 1.47 (s, 9H), 1.43 (s, 3H), 1.41 (s, 3H), 1.32 (t, J = 7.1 Hz, 3H), 0.76 (s, 9H), −0.00 (s, 3H), −0.01 (s, 3H). 13-C NMR (126 MHz; cdcl3): δ 203.7, 169.5, 164.2, 163.7, 156.0, 155.7, 152.39, 152.36, 110.0, 83.6, 83.2, 79.78, 79.71, 62.4, 61.5, 59.7, 53.2, 52.6, 52.3, 28.34, 28.31, 28.1, 26.5, 25.82, 25.78, 24.6, 18.8, 13.9, −5.1, −5.8 HRMS (FAB⁺): HRMS (FAB⁺): Calcd. for C₃₄H₆₁N₆O₁₄Si [M+H]: 805.4015; Found 805.4014.

(2R,3R)-ethyl 3-((3aR,4R,6R,6aR)-6-(benzyloxy)-2,2-dimethylt etrahydrofuro[3,4-d][1,3]dioxol-4-yl)-3-((tert-butoxycarbonyl) amino)-2-((methoxycarbonyl)amino)propanoate (18)

To a solution of compound 12 (5 g, 1 eq.) in THF (150 mL) was added 1M HCl (150 mL). The mixture was stirred 30 min. and quenched with NaHCO₃ (15 g). Methyl chloroformate (733 mg, 1 eq.) was added to the mixture and stirred an additional 2h. The layers were separated and the aqueous layer was extracted with CH₂Cl₂ (3x, 50mL) the combined extracts were washed with brine, dried over sodium sulfate and evaporated. The crude material was purified via flash chromatography (20–50% EtOAc/Hexanes) to give methyl carbamate 18 (4.07 g, 97%) as a white powder. ${[\alpha ]}_D^{25}=-96$ (c 1.04, CHCl₃) ¹H-NMR (400 MHz; CDCl₃): δ 7.38-7.27 (m, 4H), 5.61 (s, 1H), 5.25 (s, 1H), 4.90 (d, J = 12.3 Hz, 2H), 4.86 (d, J = 7.7 Hz, 1H), 4.77 (d, J = 5.9 Hz, 1H), 4.74 (d, J = 5.8 Hz, 1H), 4.60 (d, J = 12.3 Hz, 1H), 4.33 (td, J = 10.9, 2.1 Hz, 1H), 4.15-4.06 (m, 1H), 4.06-3.99 (m, 1H), 4.01 (d, J = 7.7 Hz, 1H), 3.72 (s, 3H), 1.46 (s, 3H), 1.40 (s, 9H), 1.29 (s, 3H), 1.13 (t, J = 7.1 Hz, 3H). 13-C NMR (101 MHz; cdcl3): δ 171.3, 157.3, 155.2, 137.4, 128.5, 127.78, 127.64, 112.6, 108.3, 85.73, 85.55, 82.1, 80.3, 77.4, 70.1, 62.1, 54.67, 54.56, 54.52, 52.83, 52.76, 28.3, 26.5, 25.0, 14.0. IR (Dep. CDCl₃): 3584 (br), 3362 (s), 2980(s), 2940(s), 1715(s), 1514(s). HRMS (FAB⁺): Calcd. for C₂₆H₃₉N₂O₁₀ [M+H]: 539.2605; Found 539.2602.

tert-butyl ((R)-((3aR,4R,6R,6aR)-6-(benzyloxy)-2,2-dimethylte trahydrofuro[3,4-d][1,3]dioxol-4-yl)((R)-2-oxooxazolidin-4-yl) methyl)carbamate (19)

To a solution of methyl carbamate 18 (1.08 g, 1 eq.) in THF (20 mL) was added lithium borohydride (436 mg, 10 eq.). The mixture was allowed to stir 10h or until complete by TLC. The mixture was poured over cracked ice and sat. ammonium chloride (100 mL) was added and allowed to stir for 30 min.. The mixture was extracted with EtOAc (3x, 50mL) and the combined extracts were washed with brine dried over sodium sulfate and evaporated. The crude alcohol was dissolved in MeOH (20 mL) and finely powdered KOH (1.02 g, 10 eq.) was added. The mixture was refluxed until complete by 1 hr., allowed to cool and evaporated. The crude oxazolidinone was purified via a plug of silica gel (10% MeOH/DCM) to give oxazolidinone 19 (906 mg, 98%) as a white solid. ${[\alpha ]}_D^{25}=-72$ (c 0.45, CHCl₃) ¹H-NMR (400 MHz; CDCl₃): δ 7.38-7.26 (m, 5H), 5.52 (s, 1H), 5.20 (s, 1H), 4.85 (d, J = 10.4 Hz, 1H), 4.74 (d, J = 6.0 Hz, 1H), 4.73 (d, J = 6.0 Hz, 1H), 4.65 (d, J = 12.2 Hz, 1H), 4.62 (d, J = 12.2 Hz, 1H), 4.16 (t, J = 7.6 Hz, 1H), 4.09 (t, J = 8.8 Hz, 1H), 4.00 (dd, J = 8.5, 6.2 Hz, 1H), 3.87 (d, J = 11.2 Hz, 1H), 3.74 (td, J = 10.8, 1.7 Hz, 1H), 1.46 (s, 3H), 1.45 (s, 9H), 1.31 (s, 3H). 13-C NMR (101 MHz; cdcl3): δ 160.0, 156.2, 137.0, 128.8, 128.2, 127.4, 112.8, 109.0, 86.5, 85.2, 82.4, 80.9, 70.7, 67.0, 53.1, 52.0, 28.4, 26.6, 25.0. IR (Dep. CDCl₃): 3305 (br), 2980 (s), 2938 (s), 1756 (s), 1708 (s), 1499 (s), 1455 (s). HRMS (FAB⁺): Calcd. for C₂₃H₃₃N₂O₈ [M+H]: 465.2237; Found 465.2240.

(R)-4-((R)-(2,3-bis(tert-butoxycarbonyl)guanidino)((3aR,4R,6 R,6aR)-6-(benzyloxy)-2,2-dimethyltetrahydrofuro[3,4-d][1,3] dioxol-4-yl)methyl)oxazolidin-2-one (21)

To a solution of oxazolidinone 19 (245 mg, 1eq.) and 2,6-lutidine (226 mg, 3eq.) in CH₂Cl₂ (5 mL) was added TMSOTf (352 mg, 3 eq.). The mixture was allowed to stir 1 hr. and quenched with sat. sodium bicarbonate, extracted with EtOAc (3x, 15mL). The combined extracts were dried over sodium sulfate and evaporate. The crude amine was dissolved in DMF (5mL) and triethylamine (240 mg, 4.5 eq.) and guanylating reagent 20 (153.3 mg, 1eq.) was added. The mixture was cooled to 0°C and mercuric chloride (157.5mg, 1.1 eq.) was added. The cloudy suspension was allowed to room temperature and stirred 1 hr. The reaction was quenched by adding sat. sodium chloride and ethyl acetate. After stirring the mixture for 20 min. the reaction was filtered through celite to remove all the mercury salts. The biphasic solution was separated and the organic layer was washed with H₂O (3x, 20mL), brine (20 mL), dried over sodium sulfate and evaporated in vacuo. The crude material was purified via column chromatography (20–50% EtOAc/Hexanes) to give compound 21 (216 mg, 68%). ${[\alpha ]}_D^{25}=-91$ (c 0.23, CHCl₃) ¹H-NMR (400 MHz; CDCl₃): δ 11.45 (s, 1H), 8.77 (d, J = 9.4 Hz, 1H), 7.39-7.29 (m, 5H), 5.21 (s, 1H), 5.07 (s, 1H), 4.83 (dd, J = 6.0, 1.1 Hz, 1H), 4.75 (d, J = 6.0 Hz, 1H), 4.70 (d, J = 12.4 Hz, 1H), 4.64 (d, J = 12.3 Hz, 1H), 4.50 (td, J = 10.0, 1.5 Hz, 1H), 4.17 (ddt, J = 8.7, 6.7, 2.0 Hz, 1H), 4.09 (t, J = 8.8 Hz, 1H), 4.04 (dd, J = 8.8, 6.7 Hz, 1H), 3.99 (dd, J = 10.2, 1.1 Hz, 1H), 1.49 (s, 9H), 1.49 (s, 9H), 1.47 (s, 3H), 1.32 (s, 3H). 13-C NMR (101 MHz; cdcl3): δ 163.3, 159.2, 157.2, 153.0, 136.9, 128.8, 128.2, 127.6, 113.1, 108.9, 98.7, 86.6, 85.2, 84.2, 82.0, 70.6, 66.8, 52.33, 52.17, 28.4, 28.20, 28.14, 26.7, 25.2. IR (Dep. CDCl3): 3263 (br), 2981 (s), 2936(s), 1763(s), 1725(s), 1641(s), 1613(s). HRMS (FAB⁺): Calcd. for C₂₉H₄₃N₄O₁₀ [M+H]: 607.2979; Found 607.2968.

(R)-4-((1R,2R)-1-(2,3-bis(tert-butoxycarbonyl)guanidino)-2-((4R,5S)-2,2-dimethyl-5-(((2-nitrobenzyl)amino)methyl)-1,3-dio xolan-4-yl)-2-hydroxyethyl)oxazolidin-2-one (24)

To a solution of benzyl ether 331 (300 mg) in 4:1 THF/methanol (5 mL) was added 20% Pd(OH)₂/C (150 mg). Argon was bubbled through the resulting suspension and then it was hydrogenated at 100 psi overnight. The mixture was filtered evaporated and to give lactol 22, which was taken on crude.

A solution of lactol 22, o-nitrobenzylamine HCl (456.4 mg, 5 eq.), AcOH (727 mg, 25 eq.), sodium acetate (397 mg, 10 eq.) and sodium cyanoborohydride (152 mg, 5 eq.) in MeOH (2.4 mL) was allowed to stir 1.5h. The reaction was carefully quenched with sat. sodium bicarbonate, extracted with EtOAc (3x, 20mL) and the combined extracts were dried over sodium sulfate and evaporated. The crude product wad purified via prep. TLC (10% MeOH/DCM) to give benzylamine 24 (276 mg, 86%). ${[\alpha ]}_D^{25}=-23$ (c 0.54, CHCl₃). ¹H-NMR (400 MHz; cd3cn): δ 11.59 (s, 1H), 8.72 (d, J = 9.6 Hz, 1H), 8.00 (dd, J = 8.2, 1.0 Hz, 1H), 7.69 (td, J = 7.5, 1.1 Hz, 1H), 7.58 (dd, J = 7.7, 1.2 Hz, 1H), 7.55-7.50 (m, 1H), 5.79 (s, 1H), 4.44-4.36 (m, 3H), 4.30 (ddd, J = 9.9, 5.9, 3.9 Hz, 1H), 4.18 (dd, J = 9.7, 5.7 Hz, 1H), 4.10 (dd, J = 7.2, 4.9 Hz, 1H), 4.05 (d, J = 13.5 Hz, 1H), 3.98 (d, J = 13.5 Hz, 1H), 3.75 (dd, J = 9.7, 3.7 Hz, 1H), 2.87 (dd, J = 12.1, 10.1 Hz, 1H), 2.81 (dd, J = 12.2, 3.7 Hz, 1H), 1.51 (s, 9H), 1.43 (s, 9H), 1.36 (s, 3H), 1.28 (s, 3H). 13-C NMR (101 MHz; cd3cn): δ 164.6, 159.9, 157.9, 153.8, 150.0, 134.7, 134.5, 133.2, 129.9, 125.9, 109.6, 84.3, 79.7, 78.5, 76.7, 71.6, 67.9, 53.0, 52.2, 50.7, 49.3, 28.4, 28.20, 28.18, 25.4 HRMS (FAB⁺): Calcd. for C₂₉H₄₅N₆O₁₁ [M+H]: 653.3146; Found 653.3138.

methyl N-(((4S,5R)-5-((1R,2R)-2-(2,3-bis(tert-butoxycarbonyl) guanidino)-1-hydroxy-2-((R)-2-oxooxazolidin-4-yl)ethyl)-2,2-dimethyl-1,3-dioxolan-4-yl)methyl)-N′-((4-methoxyphenyl)sul fonyl)-N-(2-nitrobenzyl)carbamimidothioate (24a)

To a solution of amine 24 (209 mg, 1 eq.) and triethylamine (124 mg, 3 eq.) in CH₂Cl₂ (4 mL, 0.1M) was added 25 (229.1 mg, 2 eq.). The mixture was allowed to stir for 45 min and evaporated. The crude compound purified via flash chromatography (1–5% MeOH/DCM) to give the isothiourea 24a (235 mg, 82%). ¹H-NMR (400 MHz; CD₃CN): δ 11.49 (s, 1H), 8.62 (d, J = 8.7 Hz, 1H), 8.00 (d, J = 8.1 Hz, 1H), 7.65 (t, J = 7.4 Hz, 1H), 7.55 (s, 1H), 7.48 (t, J = 7.7 Hz, 1H), 7.25 (d, J = 7.6 Hz, 1H), 6.87 (d, J = 8.1 Hz, 1H), 5.72 (s, 1H), 5.22 (d, J = 16.3 Hz, 1H), 5.02 (d, J = 17.6 Hz, 1H), 4.42 (t, J = 8.6 Hz, 1H), 4.41 (s, 1H), 4.31-4.27 (m, 1H), 4.22 (dd, J = 14.4, 1.8 Hz, 1H), 4.00 (dd, J = 8.5, 7.2 Hz, 1H), 3.94 (s, 1H), 3.83 (s, 3H), 3.68 (d, J = 8.8 Hz, 1H), 2.64 (s, 3H), 1.49 (s, 9H), 1.41 (s, 3H), 1.39 (s, 9H), 1.21 (s, 3H). 13-C NMR (101 MHz; cd3cn): δ 168.6, 164.3, 162.9, 159.8, 157.9, 153.7, 148.7, 137.2, 134.9, 133.0, 129.4, 129.0, 128.6, 126.2, 114.7, 110.5, 84.5, 79.8, 76.88, 76.74, 70.6, 68.0, 56.4, 54.3, 54.0, 53.0, 51.9, 28.4, 28.15, 28.05, 25.6, 19.0. (Dep. MeCN): 3317 (br), 2980 (s), 2935 (s), 1762 (s), 1727 (s), 1639 (s), 1613 (s), 1579 (s), 1529 (s). HRMS (FAB⁺): Calcd. for C₃₈H₅₄N₇O₁₄S₂ [M+H]: 896.3170; Found 896.3177.

((3aS,8R,9R,10R,10aR)-9-(2,3-(tert-butoxycarbonyl)guanidin o)-10-hydroxy-6-(((4-methoxyphenyl)sulfonyl)imino)-2,2-dim ethyl-5-(2-nitrobenzyl)octahydro-3aH-[1,3]dioxolo[4,5-e][1,3] diazonin-8-yl)methyl carbamate (27)

To a solution of compound 24a (22 mg, 1eq.) and triethylamine (22.5 mg, 5 eq.) in acetonitrile (2 mL) was added mercury (II) chloride (13.3 mg, 2 eq.). The reaction was stirred ~20 min. until the disappearance of stating material by TLC. The reaction was quenched with conc. ammonium hydroxide (2 mL) and stirred an additional 20 min. until the disappearance of stating material by TLC. The reaction mixture was filtered through celite to remove mercury salts and the was extracted 3x (5mL) with EtOAc. The combined organic layers were dried over sodium sulfate and evaporated. The crude material was purified by flash chromatography (1–5% MeOH/DCM) to give 27 as a white solid (15.1 mg, 71%). ¹H-NMR (400 MHz; CD₃CN): δ 11.48 (s, 1H), 8.64 (d, J = 8.8 Hz, 1H), 8.08 (d, J = 8.0 Hz, 1H), 7.65 (d, J = 8.2 Hz, 2H), 7.56 (t, J = 7.4 Hz, 1H), 7.48 (t, J = 7.5 Hz, 1H), 7.15 (d, J = 7.3 Hz, 1H), 6.94 (d, J = 8.7 Hz, 2H), 6.75 (s, 2H), 5.76 (s, 1H), 4.95 (s, 1H), 4.42 (t, J = 8.6 Hz, 1H), 4.36 (ddd, J = 9.7, 6.1, 1.7 Hz, 1H), 4.34-4.28 (m, 2H), 4.03 (dd, J = 9.6, 5.7 Hz, 1H), 3.99 (dd, J = 8.2, 6.7 Hz, 1H), 3.84 (s, 3H), 3.76 (d, J = 10.0 Hz, 1H), 3.40 (s, 2H), 1.49 (s, 9H), 1.45 (s, 3H), 1.39 (s, 1H), 1.22 (s, 3H). 13-C NMR (101 MHz; cd3cn): δ 164.2, 163.0, 159.8, 158.3, 157.9, 153.7, 149.0, 137.1, 134.9, 133.8, 129.3, 128.67, 128.62, 126.2, 114.7, 110.5, 84.5, 79.9, 77.7, 76.8, 70.9, 68.0, 56.3, 54.3, 52.0, 51.1, 50.2, 28.4, 28.14, 28.04, 25.5. HRMS (FAB⁺): Calcd. for C₃₇H₅₃N₈O₁₄S [M+H]: 865.3402 Found 865.3413.

((3aS,8R,9R,10R,10aR)-9-(2,3-bis(tert-butoxycarbonyl)guanid ino)-10-hydroxy-6-(((4-methoxyphenyl)sulfonyl)imino)-2,2-di methyloctahydro-3aH-[1,3]dioxolo[4,5-e][1,3]diazonin-8-yl)m ethyl carbamate (26)

Argon was bubbled through a vial containing 27 (53 mg) in THF (6 mL, wet) for 5 min. The vial was sealed and irradiated by a Hg vapor lamp for 30 min. The mixture was evaporated and purified via flash chromatography (1–5% MeOH/DCM) to give 26 (40 mg, 89%, 97% brsm) and recovered starting material (4 mg). ¹H-NMR (400 MHz; CD₃CN): δ 11.53 (s, 1H), 8.67 (d, J = 8.8 Hz, 2H), 7.75 (d, J = 8.9 Hz, 2H), 6.98 (d, J = 8.9 Hz, 2H), 6.22 (s, 2H), 5.84 (s, 1H), 4.51 (s, 1H), 4.45 (d, J = 8.6 Hz, 1H), 4.43 (d, J = 8.6 Hz, 1H), 4.37-4.35 (m, 1H), 4.30 (dt, J = 8.9, 2.7 Hz, 1H), 4.18 (dd, J = 12.5, 5.5 Hz, 1H), 4.07 (dd, J = 9.8, 5.7 Hz, 1H), 4.03 (dd, J = 8.3, 7.1 Hz, 1H), 3.83 (s, 3H), 3.78 (d, J = 10.3 Hz, 1H), 3.32 (dd, J = 7.7, 5.9 Hz, 1H), 3.28 (d, J = 5.7 Hz, 1H), 1.51 (s, 9H), 1.42 (s, 3H), 1.41 (s, 9H), 1.27 (s, 3H). HRMS (FAB⁺): Calcd. for C₃₀H₄₈N₇O₁₂S [M+H]: 730.3082; Found 730.3085

(4R,9S,10R)-4-((carbamoyloxy)methyl)-9,10-dihydroxy-6-(((4 -methoxyphenyl)sulfonyl)imino)-3,4,5,6,7,8,9,10-octahydroim idazo[4,5-e][1,3]diazonin-2(1H)-iminium 2,2,2 trifluoro-acetate (35)

To a solution of 29 (29 mg) in THF (3 mL) was added 6M HCl (3 mL). The mixture was allowed to stir for 24h. and then evaporated. The mixture was purified via RP-18 flash chromatography (100-80% 0.1% TFA in H₂O/MeCN) to give 35 (16 mg, 86%) as the TFA salt. ¹H-NMR (400 MHz; D₂O): δ 7.67 (d, J = 9.0 Hz, 2H), 6.95 (d, J = 9.0 Hz, 2H), 5.01 (s, 1H), 4.53 (t, J = 9.5 Hz, 1H), 4.44 (s, 1H), 4.14 (dd, J = 9.2, 7.2 Hz, 1H), 3.78-3.74 (m, 1H), 3.73 (s, 3H), 3.31 (d, J = 12.8 Hz, 1H), 3.11 (dd, J = 13.9, 5.8 Hz, 1H). 13-C NMR (101 MHz; D₂O): δ 162.3, 161.0, 157.3, 147.4, 133.0, 127.9, 123.5, 121.8, 114.3, 71.4, 69.5, 64.7, 55.6, 46.6, 43.7. HRMS (FAB⁺): Calcd. for C₁₇H₂₄N₇O₇S [M+H]: 470.1458 Found 470.1445.

(1S,5R,8aR)-5-((carbamoyloxy)methyl)-1-hydroxy-4,5,8,8a-tet rahydro-1H-2a,4,6,8-tetraazaacenaphthylene-3,7(2H,6H)-dii minium 2,2,2-trifluoroacetate (44)

Ketone 29 (10 mg) was cooled to 0°C and 0.25 B(TFA)₃ in TFA (1 mL) was added. The mixture was allowed to warm to r.t. slowly (leave in the ice bath) and stirred a total of 48h. The mixture was cooled to 0°C and quenched with methanol (2mL). The reaction was evaporated and diluted with methanol (3mL) and evaporated three times. The crude solid was purified via RP-18 flash chromatography (100-80% 0.1% TFA in H₂O/MeCN) to give 44 (5 mg, 92%). ¹H-NMR (500 MHz; D₂O): δ 5.26 (dd, J = 9.5, 7.3 Hz, 1H), 4.87 (d, J = 2.9 Hz, 1H), 4.77 (t, J = 9.5 Hz, 1H), 4.35 (dd, J = 9.5, 7.4 Hz, 1H), 4.22 (q, J = 3.1 Hz, 1H), 3.58 (dd, J = 13.6, 2.1 Hz, 1H), 3.38 (dd, J = 13.5, 3.8 Hz, 1H). 13-C NMR (126 MHz; d2o): δ 161.84, 161.75, 158.01, 158.01, 157.54, 157.54, 91.80, 91.80, 72.57, 72.57, 69.84, 69.84, 68.09, 68.09, 63.05, 63.05, 51.62, 51.62, 44.85, 44.85. Calcd. for C₁₀H₁₆N₇O₃ [M+H]: 282.1304 Found 282.1311.

(4R,9S)-4-((carbamoyloxy)methyl)-9-hydroxy-6-(((4-methoxy phenyl)sulfonyl)imino)-4,6,7,8,9,9a-hexahydro-1H-imidazo[4′,5′:3,4]pyrrolo[1,2-c]pyrimidin-2(3H)-iminim 2,2,2-trifluoroacetate (37)

To a flask containing compound 409 (10 mg) was added TFA (2 mL). The reaction was stirred 48 hr. and evaporated. The compound was purified via HPLC (YMC-Pack ODS-AM-322, 150x10mm, 5Å; 95-0% 0.1% TFA H₂O/MeCN over 45 min; t_r= 14.409 min.) to give 37 (9 mg, 94%) as an inseparable mixture (1:1) of diastereomers.

anti-37

¹H-NMR (500 MHz; D₂O): δ 7.80 (d, J = 9.0 Hz, 2H), 7.07 (d, J = 8.9 Hz, 2H), 5.32 (d, J = 4.1 Hz, 1H), 5.22 (dd, J = 9.2, 6.9 Hz, 1H), 4.77 (t, J = 9.4 Hz, 1H), 4.27 (dd, J = 9.2, 7.2 Hz, 1H), 4.15 (dd, J = 7.3, 3.1 Hz, 1H), 3.80 (s, 3H), 3.29 (dd, J = 14.2, 3.2 Hz, 1H), 3.16 (dd, J = 14.2, 4.1 Hz, 2H). 13-C NMR (126 MHz; d2o): δ 165.5, 162.7, 161.0, 152.0, 130.6, 125.6, 123.1, 119.1, 115.1, 69.4, 63.7, 56.0, 54.1, 46.5, 42.0

syn-37

¹H-NMR (500 MHz; D₂O): δ 7.74 (d, J = 9.0 Hz, 2H), 7.05 (d, J = 9.0 Hz, 2H), 5.49 (d, J = 5.1 Hz, 1H), 5.21 (dd, J = 9.2, 6.0 Hz, 1H), 4.63 (t, J = 9.5 Hz, 1H), 4.24 (dt, J = 7.9, 4.1 Hz, 1H), 4.17 (dd, J = 7.9, 6.8 Hz, 1H), 3.80 (s, 3H), 3.26 (dd, J = 13.0, 3.5 Hz, 1H), 2.98 (dd, J = 13.4, 7.6 Hz, 1H). 13-C NMR (126 MHz; d2o): δ 165.5, 162.7, 161.0, 152.0, 148.2, 130.6, 125.6, 123.1, 119.1, 115.1, 69.4, 63.7, 56.0, 54.1, 46.5, 42.0. HRMS (FAB⁺): Calcd. for C₁₇H₂₂N₇O₆S [M+H]: 452.1347 Found 452.1353.

(3aR,4R,9S,10R,10aR)-4-((carbamoyloxy)methyl)-9,10-dihydr oxyhexahydropyrrolo[1,2-c]purine-2,6(1H,8H)-diiminium 2,2,2-trifluoroacetate and (3R,5R,6S)-4-((amino(iminio)-methyl)amino)-3-((carbamoyloxy)methyl)-5,6-dihydroxy-2,3, 6,7-tetrahydropyrrolo[1,2-c]pyrimidin-1(5H)-iminium 2,2,2-trifluoroacetate (40)

Ketone 29 (10 mg) was cooled to 0°C and 0.25 B(TFA)₃ in TFA (1 mL) was added. The mixture was allowed to warm to r.t. slowly (leave in the ice bath) and stirred a total of 20h. The mixture was cooled to 0°C and quenched with methanol (2mL). The reaction was evaporated and diluted with methanol (3mL) and evaporated three times. The crude solid was purified via RP-18 flash chromatography (100-80% 0.1% TFA in H₂O/MeCN) to give an inseparable mixture of 40/41 (3:1, 5mg, 69%)

40

¹H-NMR (500 MHz; D₂O): δ 4.65 (dd, J = 9.7, 8.6 Hz, 1H), 4.43 (ddd, J = 8.9, 5.0, 3.8 Hz, 1H), 4.35 (dd, J = 9.5, 4.8 Hz, 1H), 4.27 (d, J = 3.5 Hz, 1H), 4.02 (ddd, J = 8.8, 6.5, 3.0 Hz, 1H), 3.69 (d, J = 8.8 Hz, 1H), 3.54 (dd, J = 14.6, 3.1 Hz, 1H), 3.44 (dd, J = 14.6, 6.6 Hz, 1H). 13-C NMR (126 MHz; d2o): δ 161.84, 161.75, 158.01, 158.01, 157.54, 157.54, 91.80, 91.80, 72.57, 72.57, 69.84, 69.84, 68.09, 68.09, 63.05, 63.05, 51.62, 51.62, 44.85, 44.85. HRMS (FAB⁺): Calcd. for C₁₀H₁₈N₇O₄ [M+H]: 300.1409 Found 300.1414.

41

¹H-NMR (500 MHz; D₂O): δ 5.28 (dd, J = 9.5, 7.0 Hz, 1H), 4.79 (t, J = 9.5 Hz, 1H), 4.66 (d, J = 7.4 Hz, 1H), 4.38 (dd, J = 9.5, 7.0 Hz, 1H), 3.96 (td, J = 7.2, 3.0 Hz, 1H), 3.52 (dd, J = 14.4, 3.0 Hz, 1H), 3.31 (dd, J = 14.5, 7.2 Hz, 1H). HRMS (FAB⁺): Calcd. for C₁₀H₁₈N₇O₄ [M+H]: 300.1409 Found 300.1414.

Figure 1.

Structures of Zetekitoxin AB (left) and saxitoxin (right)

Scheme 12.

Cyclization of hemiaminal 30

Scheme 14.

Preparation of tricycle 44

Scheme 15.

Modes of reactivity of hemiaminal 30