Total Synthesis of Pagoamide A

The first total synthesis of the thiazole-containing cyclic depsipeptide pagoamide A, is detailed. The longest linear sequence of the liquid-phase synthesis comprises 9 long linear steps from simple known starting materials, which led to the unambiguous structural confirmation of pagoamide A.


Introduction
Very recently, Gerwick and co-workers described the structure of pagoamide A [1], a thiazole-containing cyclic depsipeptide isolated from the metabolites of a marine Chlorphyte, Derbesia sp., which was collected from the shallow coastal waters near American Samoa and cultured in the laboratory. Preliminary biological tests revealed that pagoamide A had no cytotoxicity against H-460 human lung cancer cells. The isolation of this novel compound was guided by MS/MS-based molecular networking and the structure of pagoamide A (1) was elucidated through a combination of detailed 1D and 2D NMR analyses, UV, high-resolution mass spectrometry and advanced Marfey's method. Noticeably, two phenylalanine residues with opposite stereochemistry were carefully established by chemical methods, whereas the absolute configuration of the serine in the macrocyclic ring and side-chain had only been tentatively assigned by density functional theory (DFT) calculations in combination with ROESY results. As depicted in Figure 1, its unique structure comprises a 19-membered macrolactam and a pendant thiazole-bearing side chain. In total 11 amino acid moieties are incorporated in this depsipeptide with three of them possessing D configurations (D-Val, D-Ala, D-Ser). Motivated by our previous studies of 5-membered heterocycle-containing marine natural products [2][3][4][5][6][7][8][9], we hoped to accomplish the first total synthesis of pagoamide A and in turn, elucidate its relative and absolute stereochemistry. As DFT calculations for structural assignments can sometimes be error-prone [10][11][12][13], at the outset of this project, we opted for the synthesis of structure of pagoamide A (1) and one of its diastereomers 1a, where the two serine residues switched positions. The success of this investigation would allow us to further resolve the structure of this natural product. Our retrosynthetic analysis for pagoamide A (1) is presented in Scheme 1. The final step involves the connection of both the side chain 3 and the macrocycle 2. Literature precedents suggested that the ring disconnection of macrocycles, especially macrocyclic peptides, is so strategically important that it can ultimately determine the success of a synthesis [14]. We thus elected to conduct the synthesis of macrocycle 2 via two distinct routes which included macrocyclization performed by lactamization of the amino acid derived from 5 or lactonization of a seco-acid derived from 4, respectively. We envisioned the precursor 4 would be rapidly assembled through solid phase peptide synthesis (SPPS) and the linear depsipeptide 5 should be available through a more convergent solution-phase synthesis.
Molecules 2021, 26, x FOR PEER REVIEW 2 of 17 and the linear depsipeptide 5 should be available through a more convergent solutionphase synthesis.

Results
The synthetic campaign towards pagoamide A was initiated with 2-chlorotrityl chloride (CTC) resin-bound Fmoc-L-phenylalanine (6). Iterative peptide elongation was carried out under solid phase peptide synthesis conditions (Scheme 2). Fmoc-D-Val, Fmoc-D-Phe, Fmoc-D-Ala, Fmoc-O-Bn-L-Ser, Fmoc-Ot Bu-L-Thr and Fmoc-O-Bn-D-Ser were consecutively loaded onto the resin to provide the resin-immobilized heptapeptide 4. Cleavage of the resin with concomitant deprotection of the tert-butyl ether was achieved in trifluoroacetic acidtriisopropylsilanewater (95:2.5:2.5). Thus, the seco-acid was obtained in an overall 30% yield starting from CTC resin and we were thus poised to investigate the pivotal macrolactonization.
As summarized in Table 1, an exhaustive screening of conditions was conducted. Thus, seco-acid 7 was treated with the Shiina reagent (2-methyl-6-nitrobenzoic anhydride, MNBA) [15] to effect the intramolecular esterification, giving rise to the macrocycle 8 but in a very low yield (20%, entry 1).  [17], N-ethyl-N'-dimethylaminopropylcarbodiimide (EDCI), N,N'dicyclohexylcarbodiimide (DCC) [18], and N,N,N',N'-tetramethylchloroformamidinium hexafluorophosphate (TCFH) [19] all provided intractable mixtures with little or no desired products. These unsatisfactory results prompted us to explore a new approach wherein the difficult ester bond is installed at an earlier stage, and macrocyclization is performed by lactamization rather than lactonization.
Thus, condensation of Fmoc-L-Val-OH 9 and O-t Bu-L-Phe 10, mediated by (1-[bis (dimethylamino)-methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxide hexafluorophosphate) (HATU), 1-hydroxy-7-azabenzotriazole (HOAt) and N,N-diisopropylethylamine (DI-PEA) proceeded smoothly to afford dipeptide 11 in 88% yield (Scheme 3). Removal of Fmoc protecting group from the dipeptide 11 with 1,8-diazabicyclo [5.4.0]undec-7-ene (DBU) followed by coupling of the resulting amines with Fmoc-D-Phe-OH under HATU/HOAt conditions furnished tripeptide 12 in 84% yield [20]. Subsequently, functional group manipulations afforded the two desired tripeptidic acids 14 and 15 without incident. In parallel, coupling of O-All-L-Ala 16 with Fmoc-O-Bn-L-Ser-OH 17 was achieved through carboxyl activation with HATU/HOAt and provided dipeptide 18 in 98% yield. Then, applying the same sequence that was used for the synthesis of dipeptide 18, tetrapeptide 20 was prepared uneventfully from 18 through incorporation of Fmoc-O-t Bu-L-Thr and Fmoc-O-Bn-D-Ser (Scheme 3). With the tetrapeptide 20 in hand, the stage was set for the assembly of the macrocycle (Scheme 4). Esterification of the alcohol 20 with acid 14 or 15 turned out to be much more challenging than expected. A plethora of reagents was examined to mediate the coupling between alcohol 20 and acids 14 or 15. As monitored by TLC and LCMS, tetrapeptidic alcohol 20 remained intact and the acid 14/15 readily decomposed under the reaction conditions. We suspected that the formation of oxazolone upon activation of the acid might be one of the detrimental pathways. Therefore, esterification of alcohol 20 with Fmoc-L-Phe was attempted. To our delight, under Shiina conditions, ester 21 was obtained in 90% yield. Further condensation of dipeptide 22 with the free amine derived from fragment 21, mediated by HATU/HOAt gave rise to macrocyclization precursor 5 in 80% yield. Simultaneous deprotection of the allyl ester and alloc protecting groups by using Pd(PPh 3 ) 4 /ZnCl 2 /polymethylhydrosiloxane (PMHS) [21], and subsequent macrocyclization of the resulting amino acid produced macrocycle 2 in 60% yield over two steps. The final stage was set for the incorporation of bis-thiazole fragment 3 (Scheme 5), which was prepared from the known valine-derived thiazole 24 [22] via a three-step sequence including: (1) reductive amination to install the N,N-dimethyl moiety; (2) saponification in THF/H 2 O with lithium hydroxide to release the corresponding acid; (3) amidation with the other thiazole subunit 25 [23]. Trifluoroacetic acid-promoted removal of the Boc group in 2 followed by coupling of the resulting amine with the acid derived from ethyl ester 3 in the presence of EDCI/HOAt/TEA/DMAP in DMF delivered the corresponding depsipeptide 26 in 85% yield. Global deprotection was achieved by employing boron trichloride in DCM at −78 • C and furnished pagoamide A (1) in 60% yield. With many building blocks already in hand, the next step was to prepare the diastereoisomer 1a, and this was readily achieved by following the same synthetic procedure as for pagoamide A (1). The synthesis proceeded smoothly under the previous conditions through cyclization and attachment of the thiazole-containing side chain (See Supplementary Materials for details). The optical rotation of the synthetic product 1 [ α] 25 D = +8.0 (c 0.1, MeOH), was in close agreement with the value reported in the literature for natural pagoamide A [ α] 25 D = +5.5 (c 0.1, MeOH). Spectral data, including 1 H-NMR, 13 C-NMR, were collected for both the natural and synthetic sample (1) and found to be in full agreement (Table 2) [1], while the synthetic 1a showed remarkable discrepancies in spectral data as compared to natural pagoamide A, which unambiguously confirmed the original assignments of natural pagoamide A.

General Information
All reactions were conducted in flame-dried or oven-dried glassware under an atmosphere of dry nitrogen or argon. Oxygen and/or moisture-sensitive solids and liquids were transferred appropriately. The concentration of solutions in vacuo was accomplished using a rotary evaporator fitted with a water aspirator. Residual solvents were removed under a high vacuum (0.1-0.2 mm Hg). All reaction solvents were purified before use: tetrahydrofuran (THF) was distilled from Na/benzophenone. Toluene was distilled over molten sodium metal. Dichloromethane (DCM), 1,2-dichloroethane (DCE) and trimethylamine (Et 3 N) were distilled from CaH 2 . Methanol (MeOH) was distilled from Mg/I 2 . The reagents were purchased at the highest commercial quality and used without further purification unless otherwise stated. Flash column chromatography was performed using the indicated solvents on silica gel 60 (230-400 mesh ASTM E. Qingdao, Tsingtao, China). Reactions were monitored using thin-layer chromatography (TLC), which was carried out using pre-coated sheets (Qingdao silica gel 60-F250, 0.2 mm). Compounds were visualised with UV light, iodine and ceric ammonium molybdate stainer phosphomolybdic acid in EtOH. The 1 H NMR spectra were recorded on Avance 300 MHz, Avance 400 MHz or Avance 500 MHz spectrometers (Bruker, Karlsruhe, Germany). Chemical shifts were reported in parts per million (ppm), relative to either a tetramethylsilane (TMS) internal standard or the signals due to the solvent. The following abbreviations are used to describe the spin multiplicity: s = singlet, d = doublet, t = triplet, q = quartet, qn = quintet, m = multiplet, br = broad, dd = doublet of doublets, dt = doublet of triplets, dq = doublet of quartets, ddd = doublet of doublet of doublets; other combinations are derived from those listed above. Coupling constants (J) are reported in Hertz (Hz) for corresponding solutions, and chemical shifts are reported as parts per million (ppm) relative to residual CHCl 3 δH (7.26 ppm). 13 C-NMR nuclear magnetic resonance spectra were recorded at 75 MHz, 100 MHz or 125 MHz for corresponding solutions, and chemical shifts are reported as parts per million (ppm) relative to residual CDCl 3 δC (77.16 ppm). High-resolution mass spectra were measured on an ABI Q-star Elite (Beijing, China). Optical rotations were recorded on a Rudolph AutoPol-I polarimeter (Shanghai, China) at 589 nm with a 50 mm cell. Data are reported as follows: specific rotation (c (g/100 mL), solvent).

General Experimental Procedures
The linear peptides were prepared by manual SPPS using 2-chlorotrityl resin (GL Biochem, loading: ca. 1.0 mmol/g) and the general procedure was stated as follows: Anhydrous DCM was employed to swell the resin in a suitable syringe for half an hour. Subsequently, the swollen resin was washed with DCM (3 × 30 mL), and a solution of Fmoc-L-Phe (5.0 equiv. to the resin capacity) and DIPEA (10.0 equiv. to the resin capacity) in DMF was added to this reaction vessel. The reaction mixture was shaken at room temperature for 8 h. After that, the resin was washed with DCM (3 × 30 mL) and DMF (3 × 30 mL), respectively. Next, the resin underwent iterative peptide assembly (Fmoc-SPPS). A solution of Fmoc-protected amino acid (4.0 equiv. according to the resin capacity), HATU (4.0 equiv.) and DIPEA (8.0 equiv.) in DMF was gently agitated with the resin at room temperature for 2 h. Subsequently, the resin was washed with DCM (3 × 30 mL) and DMF (3 × 30 mL). After Kaiser test showed complete coupling, the deprotection of Fmoc group at N-terminus was achieved by 30 min treatment of 50% morpholine in DMF at room temperature. As followed, the resin was washed thoroughly with DCM (3 × 30 mL) and DMF (3 × 30 mL). The coupling and deprotecting procedures were repeated for coupling each amino acid until the desired linear peptide sequence 4 was obtained.

((Allyloxy)carbonyl)-D-phenylalanyl-L-valine (22)
Step A: To a solution of Alloc-D-Phe (7.5 g, 30.3 mmol, 1.0 eq.), l-valine methyl ester hydrochloride (10.2 g, 60.6 mmol, 2.0 eq.), HOAt (8.2 g, 60.6 mmol, 2.0 eq.) and DIPEA (20 mL, 121.2 mmol, 4.0 eq.) in dry DCM (200 mL, 0. 15 M) under an argon atmosphere, was added EDCI (11.6 mg, 60.6 mmol, 2.0 eq.) at 0 • C. The mixture was allowed to stir for 12 h at room temperature and then concentrated in vacuo furnishing a solid residue. The solid residue was redissolved in EtOAc (5 mL) and quenched with 4% citric acid aqueous solution. The aqueous layer was extracted with EtOAc (3 × 10 mL). The combined organic layers were washed with saturated aqueous solution of NaHCO 3 (100 mL), brine (100 mL), dried over anhydrous Na 2 SO 4 , filtered and concentrated in vacuo. Purification of the crude product was performed by flash chromatography on silica gel (hexanes/EtOAc = 3/1) to afford S1 (9. were added at 0 • C, and the mixture was stirred for 5 min at the same temperature. Then, HATU (1.1 g, 2.8 mmol, 2.0 eq.) was added at 0 • C and the resulting mixture was allowed to stir for 12 h at room temperature. The reaction mixture was quenched by the addition of H 2 O (30 mL). Layers were separated and the aqueous phase was extracted with EtOAc (3 × 20 mL). The combined organic layers were washed with brine (20 mL), dried over anhydrous Na 2 SO 4 , filtered and concentrated in vacuo. Purification of the crude product was performed by flash chromatography on silica gel (hexanes/EtOAc = 2/3) to afford 5 (1. atmosphere at 0 • C. The mixture was allowed to stir for 12 h at room temperature and then concentrated in vacuo furnishing a solid residue. The solid residue was redissolved in EtOAc (20 mL) and quenched with 4% citric acid aqueous solution. The aqueous layer was extracted with EtOAc (3 × 20 mL). The combined organic layers were washed with saturated aqueous solution of NaHCO 3 (20 mL), brine (20 mL), dried over anhydrous Na 2 SO 4 , filtered and concentrated in vacuo. Purification of the crude product was performed by flash chromatography on silica gel (hexanes/EtOAc = 2/1) to afford 2 (397 mg, 60% for 2 steps) as a white solid. TLC: R f = 0.5 (hexanes/EtOAc = 1/1), UV & PMA stain. α 27 D = +17.9 (c 7.9, CHCl 3 ). 1  Step A: To a solution of 2 (180 mg, 0.18 mmol, 1.0 eq.) in DCM (3 mL, 0.05 M) was added TFA (1 mL) dropwise at 0 • C. After being stirred at room temperature for 3 h, the reaction mixture was concentrated in vacuo to afford the crude amine 3, which was used directly in the next step without further purification. To a solution of 3 (285.0 mg, 0.72 mmol, 4.0 eq.) in THF/H 2 O (4 mL/2 mL, 0.12 M) was added LiOH·H 2 O (92.0 mg, 2.2 mmol, 12.0 eq.) at 0 • C. After being stirred for 3 h at room temperature, the reaction mixture was quenched with concentrated HCl (1 mL), and all solvent was removed in vacuo providing the crude acid which was used directly in the next step without further purification. To a solution of the above crude amine (0.18 mmol, 1.0 eq.) and the crude acid (0.72 mmol, 4.0 eq.) in DMF (5 mL, 0.09 M) were added HOAt (98 mg, 0.72 mmol, 4.0 eq.), Et 3 N (0.2 mL, 1.4 mmol, 8.0 eq.), DMAP (11.0 mg, 0.04 mmol, 0.5 eq.) and EDCI (138.0 mg, 0.72 mmol, 4.0 eq.) under argon atmosphere at 0 • C. The mixture was allowed to stir for 36 h at room temperature and then concentrated in vacuo furnishing a solid residue. The solid residue was redissolved in EtOAc (10 mL) and quenched with saturated aqueous solution of NaHCO 3 (10 mL). The aqueous layer was extracted with EtOAc (3 × 10 mL). The combined organic layers were washed with brine (10 mL), dried over anhydrous Na 2

Conclusions
In summary, the first total synthesis of the depsipeptide pagoamide A (1), was accomplished efficiently. Obstacles were circumvented by judicious synthetic design that featured a sequence-dependent esterification via Shiina conditions. The route entails a longest sequence of nine linear steps to access pagoamide A (1) starting from 17, which served to support the unequivocal structural elucidation of this natural product.