Total Synthesis of the Highly N-Methylated Peptides Carmabin A and Dragomabin

The first total synthesis of carmabin A and dragomabin was achieved at 52.3 mg and 43.8 mg scale, respectively. The synthesis led to determination of the configuration of carmabin A and reassignment of the configuration of dragomabin at the stereogenic centre on the alkyne-bearing fragment.


Introduction
Secondary metabolites produced by marine microorganisms are a rich source of valuable pharmaceuticals ranging from antimalarial agents to anticancer agents [1][2][3][4]. Marine cyanobacteria are widely distributed throughout the world [5] and produce a large number of structurally complex secondary metabolites containing alkynyl groups [6]. Due to their intriguing biological activities, these compounds have attracted the attention of many synthetic chemists [7][8][9][10][11][12][13], including us [14]. Carmabin A and dragomabin, two acetylene-containing lipopeptides, were discovered in 1998 [15,16] and 2007 [16] as secondary metabolites of cyanobacterium Lyngbyamajuscula in Panama by Gerwick et al. Structurally characterized with multiple N-methylated amino acids and a lipid chain (Mdya for carmabin A, Moya for dragomabin), carmabin A showed antimalarial activity (4.3 µM) towards the W2 chloroquine-resistant malaria strain and inhibitory activity towards mammalian Vero cells. Dragomabin, which has lost a propyl unit compared to carmabin A, showed antimalarial activity (6.0 µM) towards the W2 chloroquine-resistant malaria strain but very weak cytotoxicity to Vero cells. Thus, dragomabin possesses significant differential toxicity between parasite and mammalian cells. However, due to the scarceness of carmabin A and dragomabin, the medicinal chemistry as well as the mechanism of action have not been explored.
Without sufficient material, Gerwick et al. could not degrade carmabin A to obtain free Mdya to determine the configuration of the stereogenic centre by comparison with synthetic isomers [16]. They proposed a configuration of 35S, 37R or 35R, 37S for the Mdya fragment of carmabin A according to the strong correlations between H-35 and CH 3 -45 and between H-35 and CH 3 -33 [16]. The absolute stereochemistry of dragomabin was not clearly elucidated. Gerwick et al. proposed that dragomabin possesses a configuration of 35S for the Moya fragment according to the comparison of spectroscopic data between dragomabin and 35S-dragonamide A and 35R-dragonamide A [16].
In addition, the highly N-methylated amino acids of carmabin A and dragomabin led to two or more discernible conformers, which complicated the determination of absolute configuration (e.g., carmabin A: five conformers with a ratio of 23:6:5:1:1 in DMSO-d 6 1 H NMR spectrum) [15].
Synthetic access to useful quantities of carmabin A and dragomabin has been hampered by their unknown stereochemical configurations and conformers caused by the N-methyl groups, which further undermines investigation of those lipopeptides as potential lead compounds as well as the study of their mechanism of action. Further medicinal research enabled by chemical synthesis of carmabin A and dragomabin may provide new leads for drug discovery. Herein, we report the first total synthesis and absolute stereochemical determination of carmabin A at C35, C37. In addition, we report the first total synthesis of dragomabin with a revision of the stereochemical configuration at C35.

Results
Scheme 1 illustrates our retrosynthetic analysis. The C-terminal amide of carmabin A and dragomabin could be prepared via amidation of the C-terminal methyl esters of compounds 3 and 4. Compounds 3 and 4 were further disconnected into two parts: the Mdya/Moya fragment and protected tetrapeptide 7. Tetrapeptide 7 could be prepared by repeated condensation reactions of amino acids. The methyl group of Mdya/Moya 5/6 could be stereoselectively installed following the methodology developed by Evans [17][18][19].
Mar. Drugs 2018, 16, x FOR PEER REVIEW 2 of 11 carmabin A: five conformers with a ratio of 23:6:5:1:1 in DMSO-d6 1 H NMR spectrum) [15]. Synthetic access to useful quantities of carmabin A and dragomabin has been hampered by their unknown stereochemical configurations and conformers caused by the N-methyl groups, which further undermines investigation of those lipopeptides as potential lead compounds as well as the study of their mechanism of action. Further medicinal research enabled by chemical synthesis of carmabin A and dragomabin may provide new leads for drug discovery. Herein, we report the first total synthesis and absolute stereochemical determination of carmabin A at C35, C37. In addition, we report the first total synthesis of dragomabin with a revision of the stereochemical configuration at C35.

Results
Scheme 1 illustrates our retrosynthetic analysis. The C-terminal amide of carmabin A and dragomabin could be prepared via amidation of the C-terminal methyl esters of compounds 3 and 4. Compounds 3 and 4 were further disconnected into two parts: the Mdya/Moya fragment and protected tetrapeptide 7. Tetrapeptide 7 could be prepared by repeated condensation reactions of amino acids. The methyl group of Mdya/Moya 5/6 could be stereoselectively installed following the methodology developed by Evans [17][18][19]. Scheme 1. Retrosynthetic analysis of carmabin A (1) and dragomabin (2).

Synthesis of Mdya 5 of Carmabin A
The forward synthesis started with the synthesis of acid 5 (Scheme 2). Carboxylic acid 8 was prepared according to the literature. By employing the route we previously developed for the synthesis of ent-5 [14], 8 was converted to corresponding alcohol 12 in a four-step sequence involving acyl chlorination, amidation with benzyl-2-oxazolidinone, diastereoselective α-methylation, and reduction in 46% overall yield (four steps). Alcohol 12 was converted to iodide 13 under Appel reaction conditions [20]. Iodide 13 was then subjected to asymmetric alkylation with the enolate generated from 14 to afford 15 [17] in 66% yield (d.r. > 20:1 according to 1 H NMR), which constructed Scheme 1. Retrosynthetic analysis of carmabin A (1) and dragomabin (2).

Synthesis of Mdya 5 of Carmabin A
The forward synthesis started with the synthesis of acid 5 (Scheme 2). Carboxylic acid 8 was prepared according to the literature. By employing the route we previously developed for the synthesis of ent-5 [14], 8 was converted to corresponding alcohol 12 in a four-step sequence involving acyl chlorination, amidation with benzyl-2-oxazolidinone, diastereoselective α-methylation, and reduction in 46% overall yield (four steps). Alcohol 12 was converted to iodide 13 under Appel reaction conditions [20]. Iodide 13 was then subjected to asymmetric alkylation with the enolate generated from 14 to afford 15 [17] in 66% yield (d.r. > 20:1 according to 1 H NMR), which constructed the 2S,4R stereochemistry on the Mdya fragment. Treatment of resulting 15 with 1 N HCl led to the hydrolysis of both the TMS group and the amide bond, providing 5 in 61% yield (2.1 g scale, one batch).

Synthesis of Tetrapeptide 7
Scheme 3 depicts the construction of tetrapeptide 7. Boc deprotection of 16 using TFA followed by condensation with 17 under the conventional coupling conditions (HATU/DIPEA) discovered by Louis A. Carpino [21] produced 18 in 85% yield. By repeating the condensation under the same conditions, Boc-L-Ala 19 and Boc-N-Me-L-Phe 21 were sequentially introduced, providing protected tetrapeptide 7. The relatively low yield of the condensation can probably be attributed to the steric hindrance imposed by the additional methyl group. Problematic undesired diketopiperazine (DKP) formation [22,23] was not observed during the condensation reaction.

Total Synthesis of Carmabin A and Discussion
With 7, 5, ent-5 in hand, we next aimed at completing the total synthesis of carmabin A and confirming its absolute stereochemistry (Scheme 4). The final steps involved coupling reactions between Boc-deprotected 7 with 5 and with ent-5 followed by treatment with ammonia to produce 1 and 1a. After deprotection of 7 under 4 M HCl, the resulting intermediate was subjected to coupling reaction directly, while only a trace amount of condensation product was observed. This could be explained by that 7 or Boc-deprotected 7 are unstable under 4 M HCl. By switching 4 M HCl for 10% TFA, the coupling reaction proceeded smoothly to provide 3 and 3a in 61% and 59% yields, respectively. The transformation from 3 to 1 was first conducted in anhydrous NH3 solution (7.0 M Scheme 2. Synthesis of Mdya 5.

Synthesis of Tetrapeptide 7
Scheme 3 depicts the construction of tetrapeptide 7. Boc deprotection of 16 using TFA followed by condensation with 17 under the conventional coupling conditions (HATU/DIPEA) discovered by Louis A. Carpino [21] produced 18 in 85% yield. By repeating the condensation under the same conditions, Boc-L-Ala 19 and Boc-N-Me-L-Phe 21 were sequentially introduced, providing protected tetrapeptide 7. The relatively low yield of the condensation can probably be attributed to the steric hindrance imposed by the additional methyl group. Problematic undesired diketopiperazine (DKP) formation [22,23] was not observed during the condensation reaction.

Synthesis of Tetrapeptide 7
Scheme 3 depicts the construction of tetrapeptide 7. Boc deprotection of 16 using TFA followed by condensation with 17 under the conventional coupling conditions (HATU/DIPEA) discovered by Louis A. Carpino [21] produced 18 in 85% yield. By repeating the condensation under the same conditions, Boc-L-Ala 19 and Boc-N-Me-L-Phe 21 were sequentially introduced, providing protected tetrapeptide 7. The relatively low yield of the condensation can probably be attributed to the steric hindrance imposed by the additional methyl group. Problematic undesired diketopiperazine (DKP) formation [22,23] was not observed during the condensation reaction.

Total Synthesis of Carmabin A and Discussion
With 7, 5, ent-5 in hand, we next aimed at completing the total synthesis of carmabin A and confirming its absolute stereochemistry (Scheme 4). The final steps involved coupling reactions between Boc-deprotected 7 with 5 and with ent-5 followed by treatment with ammonia to produce 1 and 1a. After deprotection of 7 under 4 M HCl, the resulting intermediate was subjected to coupling reaction directly, while only a trace amount of condensation product was observed. This could be explained by that 7 or Boc-deprotected 7 are unstable under 4 M HCl. By switching 4 M HCl for 10% TFA, the coupling reaction proceeded smoothly to provide 3 and 3a in 61% and 59% yields, respectively. The transformation from 3 to 1 was first conducted in anhydrous NH3 solution (7.0 M Scheme 3. Synthesis of tetrapeptide 7.

Total Synthesis of Carmabin A and Discussion
With 7, 5, ent-5 in hand, we next aimed at completing the total synthesis of carmabin A and confirming its absolute stereochemistry (Scheme 4). The final steps involved coupling reactions between Boc-deprotected 7 with 5 and with ent-5 followed by treatment with ammonia to produce 1 and 1a. After deprotection of 7 under 4 M HCl, the resulting intermediate was subjected to coupling reaction directly, while only a trace amount of condensation product was observed. This could be explained by that 7 or Boc-deprotected 7 are unstable under 4 M HCl. By switching 4 M HCl for 10% TFA, the coupling reaction proceeded smoothly to provide 3 and 3a in 61% and 59% yields, respectively. The transformation from 3 to 1 was first conducted in anhydrous NH 3 solution (7.0 M in MeOH) at room temperature. With most of starting material left in the reaction system, only 10% yield of 1 was observed. Elevating the temperature to 40 • C improved the yield to 40%. When the reaction was transferred to a sealed tube in 60 • C, 1 and 1a were obtained in 67% and 64% yield, respectively. After careful comparison, the spectroscopic data of synthetic 1 were consistent within the margin of error (0.02 ppm for 1 H NMR and 0.2 ppm for 13 C NMR) for the data originally reported for natural 1. The NMR data of synthetic 1a did not match the data of natural 1. Consequently, the absolute configuration of carmabin A was established as shown in 1.

Mar. Drugs 2018, 16, x FOR PEER REVIEW 4 of 11
in MeOH) at room temperature. With most of starting material left in the reaction system, only 10% yield of 1 was observed. Elevating the temperature to 40 °C improved the yield to 40%. When the reaction was transferred to a sealed tube in 60 °C, 1 and 1a were obtained in 67% and 64% yield, respectively. After careful comparison, the spectroscopic data of synthetic 1 were consistent within the margin of error (0.02 ppm for 1 H NMR and 0.2 ppm for 13 C NMR) for the data originally reported for natural 1. The NMR data of synthetic 1a did not match the data of natural 1. Consequently, the absolute configuration of carmabin A was established as shown in 1.

Synthesis of Moya 6 of Dragomabin
Next, we turned our attention to dragomabin. The Moya 6 fragment of dragomabin was obtained from ent-11 via deprotection of TMS with TFA followed by removal of the chiral auxiliary (Scheme 5).

Total Synthesis of Dragomabin and Discussion
After deprotection of 7 under 4 M HCl, the resulting intermediate was subjected to a coupling reaction, which proceeded smoothly to provide 4 in 55% yield. Amidation of 4 with anhydrous NH3 solution afforded compound 2 in 61% yield (Scheme 6). Compound 2 was prepared with the same stereochemistry as that reported in the isolation paper. However, neither the 1 HNMR data nor the optical rotation data are consistent with the reported values. The main concern regarding the true structure was the configuration at C35. By employing the same procedure described above, ent-6 was obtained from 11 and then subjected to condensation with tetrapeptide 7 to give 2a. It was indeed found that the data for 2a were consistent with those of natural dragomabin. Thus, the correct structure for dragomabin was revised as shown in 2a. As dragomabin and dragonamide differ in the stereochemistry on the Moya fragment, it appears that the stereochemistry on the alkyne fragment in these lipopeptides is variable and correlation with other natural products is not reliable.

Synthesis of Moya 6 of Dragomabin
Next, we turned our attention to dragomabin. The Moya 6 fragment of dragomabin was obtained from ent-11 via deprotection of TMS with TFA followed by removal of the chiral auxiliary (Scheme 5).
Mar. Drugs 2018, 16, x FOR PEER REVIEW 4 of 11 in MeOH) at room temperature. With most of starting material left in the reaction system, only 10% yield of 1 was observed. Elevating the temperature to 40 °C improved the yield to 40%. When the reaction was transferred to a sealed tube in 60 °C, 1 and 1a were obtained in 67% and 64% yield, respectively. After careful comparison, the spectroscopic data of synthetic 1 were consistent within the margin of error (0.02 ppm for 1 H NMR and 0.2 ppm for 13 C NMR) for the data originally reported for natural 1. The NMR data of synthetic 1a did not match the data of natural 1. Consequently, the absolute configuration of carmabin A was established as shown in 1.

Synthesis of Moya 6 of Dragomabin
Next, we turned our attention to dragomabin. The Moya 6 fragment of dragomabin was obtained from ent-11 via deprotection of TMS with TFA followed by removal of the chiral auxiliary (Scheme 5).

Total Synthesis of Dragomabin and Discussion
After deprotection of 7 under 4 M HCl, the resulting intermediate was subjected to a coupling reaction, which proceeded smoothly to provide 4 in 55% yield. Amidation of 4 with anhydrous NH3 solution afforded compound 2 in 61% yield (Scheme 6). Compound 2 was prepared with the same stereochemistry as that reported in the isolation paper. However, neither the 1 HNMR data nor the optical rotation data are consistent with the reported values. The main concern regarding the true structure was the configuration at C35. By employing the same procedure described above, ent-6 was obtained from 11 and then subjected to condensation with tetrapeptide 7 to give 2a. It was indeed found that the data for 2a were consistent with those of natural dragomabin. Thus, the correct structure for dragomabin was revised as shown in 2a. As dragomabin and dragonamide differ in the stereochemistry on the Moya fragment, it appears that the stereochemistry on the alkyne fragment in these lipopeptides is variable and correlation with other natural products is not reliable.

Total Synthesis of Dragomabin and Discussion
After deprotection of 7 under 4 M HCl, the resulting intermediate was subjected to a coupling reaction, which proceeded smoothly to provide 4 in 55% yield. Amidation of 4 with anhydrous NH 3 solution afforded compound 2 in 61% yield (Scheme 6). Compound 2 was prepared with the same stereochemistry as that reported in the isolation paper. However, neither the 1 HNMR data nor the optical rotation data are consistent with the reported values. The main concern regarding the true structure was the configuration at C35. By employing the same procedure described above, ent-6 was obtained from 11 and then subjected to condensation with tetrapeptide 7 to give 2a. It was indeed found that the data for 2a were consistent with those of natural dragomabin. Thus, the correct structure for dragomabin was revised as shown in 2a. As dragomabin and dragonamide differ in the stereochemistry on the Moya fragment, it appears that the stereochemistry on the alkyne fragment in these lipopeptides is variable and correlation with other natural products is not reliable.
All reactions were carried out under an argon atmosphere with dry solvents under anhydrous conditions, unless otherwise noted. Tetrahydrofuran (THF) was distilled immediately before use from sodium-benzophenone ketyl. Diisopropylamine was distilled from calcium hydride. Solvents for chromatography were used as supplied by Tianjin Reagents Chemical (Tianjin, China). Reactions were monitored by thin-layer chromatography (TLC) carried out on silica gel plates, using UV light as the visualizing agent and aqueous phosphomolybdic acid or basic aqueous potassium permanganate as the developing agent. A 200-300 mesh silica gel was used for column chromatography.
Optical rotations were recorded on an Insmark IP 120 digital polarimeter (Insmark, Shanghai, China). IR spectra were recorded on a Bruker Tensor 27 instrument (Ettlingen, Germany). Only the strongest and/or most structurally important absorptions of IR spectra were reported in wavenumbers (cm −1 ). 1 H NMR, 13 C NMR, and 2D NMR were recorded on Bruker AV 400 and calibrated by using internal references and solvent signals CDCl3 (δH = 7.26 ppm, δC = 77.16 ppm), unless otherwise noted. 1 H NMR data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, p = quintet, br = broad, m = multiplet), coupling constants and integration. High-resolution mass spectra (HRMS) were detected on an IonSpec FT-ICR mass spectrometer by Varian 7.0T FTMS (Kuala Lumpur, Malaysia).
All reactions were carried out under an argon atmosphere with dry solvents under anhydrous conditions, unless otherwise noted. Tetrahydrofuran (THF) was distilled immediately before use from sodium-benzophenone ketyl. Diisopropylamine was distilled from calcium hydride. Solvents for chromatography were used as supplied by Tianjin Reagents Chemical (Tianjin, China). Reactions were monitored by thin-layer chromatography (TLC) carried out on silica gel plates, using UV light as the visualizing agent and aqueous phosphomolybdic acid or basic aqueous potassium permanganate as the developing agent. A 200-300 mesh silica gel was used for column chromatography.
Optical rotations were recorded on an Insmark IP 120 digital polarimeter (Insmark, Shanghai, China). IR spectra were recorded on a Bruker Tensor 27 instrument (Ettlingen, Germany). Only the strongest and/or most structurally important absorptions of IR spectra were reported in wavenumbers (cm −1 ). 1 H NMR, 13 C NMR, and 2D NMR were recorded on Bruker AV 400 and calibrated by using internal references and solvent signals CDCl 3 (δ H = 7.26 ppm, δ C = 77.16 ppm), unless otherwise noted. 1 H NMR data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, p = quintet, br = broad, m = multiplet), coupling constants and integration. High-resolution mass spectra (HRMS) were detected on an IonSpec FT-ICR mass spectrometer by Varian 7.0T FTMS (Kuala Lumpur, Malaysia).

Conclusions
In summary, carmabin A and dragomabin have been synthesized in an efficient and stereoselective fashion (52.3 mg and 43.8 mg obtained, respectively). The absolute stereochemistry at C35 and C37 of carmabin A has been assigned as 35R, 37S and the absolute stereochemistry at C35 of dragomabin has been revised as 35R. It is anticipated that this work will lead to further investigations of carmabin A and dragomabin as well as other acetylene-containing lipopeptides in both medicinal chemistry and chemical biology. Funding: We acknowledge financial support from Fundamental Research Funds for the Central Universities, the National Natural Science Foundation of China (NSFC) (81573282 to Y.C. and 81703350 to L.W.), the National Science Fund for Distinguished Young Scholars (81625021 to Y.C.), and the Hundred Young Academic Leaders Program of Nankai University.