Synthesis of Chondramide A Analogues with Modified β-Tyrosine and Their Biological Evaluation

Abstract

Starting from cinnamates 9, obtained by Wittig reaction or Heck coupling, the diols 17 were prepared by asymmetric dihydroxylation. This was followed by a regioselective substitution of the 3-OH group with hydrazoic acid under Mitsunobu conditions. Methylation of the 2-OH group and reduction of the azide group led to the β-tyrosine derivatives 8. Condensation with the dipeptide acid 6 furnished the tripeptide part of the chondramides. The derived acids 21 were combined with the hydroxy ester 7 to the esters 22. Cleavage of the tert-butyl groups and intramolecular lactam formation gave rise to the chondramide A analogues 2b–k. Growth inhibition assays showed most of the analogues to be biologically active. Some of them even reach the activity of jasplakinolide. It can be concluded that the 4-position of the aryl ring in the β-tyrosine of chondramide A tolerates structural modifications quite well.

Introduction

Cyclodepsipeptides comprise a unique class of secondary metabolides. They frequently contain unusual amino acids, like D-amino acids or N-methylated amino acids and hydroxy acids that typically originate from the polyketide pathway. Incorporation of the hydroxy acid results in an ester bond, explaining the term “depsi”.^1,2 Many cyclodepsipeptides have been isolated from marine sponges and found to display interesting biological activities.³ Thus, cyclodepsipeptides with anti-HIV or anti-tumor activity are known. With the peptide subunit cyclodepsipeptides are clearly protein-like and therefore it is not surprising that they very often modulate protein-protein interactions.⁴

Very prominent examples of such cyclodepsipeptides, displaying anti-tumor activity are the jasplakinolides and the chondramides. Jasplakinolide (1) was isolated from the marine sponge Jaspis splendans many years ago (Figure 1).⁵ Independently, it was also found in a Jaspis sponge and named Jaspamide.⁶ In the meantime jasplakinolide was isolated from many other sponges. Furthermore, several additional jasplakinolides could be isolated by the groups of Zampella⁷ and Crews.⁸ In contrast, so far only four natural chondramides are known, namely chondramides A-D (2–5) (Figure 1). They were isolated by the Höfle/Reichenbach group from myxobacteria.⁹ The chondramides are quite similar in structure to jasplakinolide, in that they contain a tripeptide subunit consisting of an L-alanine, a N-methyl-D-tryptophan, and an L-β-tyrosine. Jasplakinolide and the chondramides differ essentially in the hydroxy acid that bridges the tripeptide part to form the cyclodepsipeptide. Whereas jasplakinolide contains an 8-hydroxy acid, in the chondramides the corresponding sector is a 7-hydroxy acid. Accordingly, jasplakinolide features a 19-membered macrocycle, while the chondramides are 18-membered.

Figure 1.

Structures of jasplakinolide (1) and the chondramides A–D (2–5).

Over the years several total syntheses of jasplakinolide have been achieved. ¹⁰ Most of them close the macrocyclic ring by intramolecular amide or ester formation. The Waldmann/Arndt synthesis^10g relies on a relay ring-closing metathesis reaction. Furthermore, syntheses of simplified jasplakinolide analogues have been reported.^11,12 Surprisingly, syntheses of chondramides were achieved only recently. The reason was that the configuration at the stereogenic centers was not given in the original article. Even though it could be assumed that the configurations in the peptide part and even the overlapping hydroxy acid might be similar to the ones in jasplakinolide. The Waldmann ¹³ and Kalesse ¹⁴ groups independently secured the configuration of the hydroxy acid in the chondramides by synthesizing various stereoisomers of chondramide C. Our group had developed a concise synthesis of the hydroxy acid through an asymmetric vinylogous aldol reaction.¹⁵ Subsequently, we reported the total synthesis of chondramide A (chon A).¹⁶ Instead of a β-tyrosine, this chondramide contains a 3-amino-2-methoxy-3-arylpropanoic acid, whose configuration was determined to be (2S,3S). The ester bond could be established either by Yamaguchi esterification or via a Mitsunobu reaction¹⁷ that was then followed by macrolactam formation. While the β-tyrosine derivative in chondramide A seems more complicated on first sight, it is easily available by asymmetric dihydroxylation of a cinnamic ester precursor.

The potent antitumor activity of these cyclodepsipeptides is due to their stabilizing effect on F-actin filaments. ¹⁸ Because of this property jasplakinolide became an important tool in cell biology. Together with microtubules and the intermediate filaments, actin filaments (F-actin) make up the cytoskeleton which has a key role in cell shape and division. According to a binding model recently refined by the Waldmann group, chondramide C is located in a shallow binding pocket made up by three independent actin subunits in the filament.¹⁹ The binding site is identical to the one of the bicyclic heptapeptide phalloidin.²⁰ There seems to be a hydrophobic interaction of the indole moiety with a loop region of one of the actin subunits (subunit X). The phenolic group of the β-tyrosine occupies a larger cavity with the hydroxyl group close to an Asp of G-actin monomer Y. It seems that the hydroxy acid not only serves to provide a scaffolding role, but rather parts (C1-C3) of it are close to the protein surface of subunit Z and contribute substantially to efficient binding. This model was supported by the activity of the available jasplakinolide and chondramide analogues (Figure 2).¹² For example, omitting the phenolic ring essentially results in an inactive jasplakinolide analogue. With regard to the hydroxy acid, it turned out that the configuration at C2 of the methyl-bearing carbon and the E-configuration of the C4-C5 double bond are crucial. Jasplakinolide analogues with this feature are still active even if they lack some of the other methyl groups of the hydroxy acid. For chondramide C analogues it could be shown that the configuration at C6 and C7 of the hydroxy acid is less critical.

Figure 2.

Key structure-activity relationships for good biological activity of jasplakinolides and chondramides illustrated with the jasplakinolide structure.

Due to their interesting biology and structural similarity analogues of the above-mentioned cyclodepsipeptides should help to understand protein-protein interactions and further advance the binding site model.¹² With an efficient route to chondramide A we embarked on the synthesis of chondramide A analogues with modified β-tyrosine derivatives. In this paper we report on the synthesis of ten analogues and their biological activity.

Results and Discussion

Synthesis of the Analogues

The synthetic route towards natural chondramide A (2a), recently disclosed from our laboratory,¹⁶ was put as a basis for synthesis of new chondramide analogues, containing variations in the β-tyrosine unit.¹⁶ Accordingly, this implied the use of differently substituted β-amino esters 8 that had to be synthesized from substituted trans-cinnamic esters 9. At this point, taking into account the compatibility of the substituents at the aromatic ring with all the synthetic steps, we have chosen eight cinnamic esters 9b–e, 9g–i, and 9k as precursors for the desired chondramide analogues (Scheme 2).

Scheme 2.

Synthesis of cinnamic acid building blocks 9 by Wittig reaction or Heck coupling; NMP = N-methyl-2-pyrrolidone.

All the trans-cinnamic esters 9 were synthesized from easily accessible commercial starting materials as outlined in Scheme 2. Enoates 9b–e, 9k were prepared by Wittig reaction of the corresponding aldehydes 10 with the stabilized ylide [(carbomethoxy)methylene]triphenylphosphorane ²¹ in dichloro-methane. Enoate 9e could also be obtained by Knoevenagel condensation with subsequent esterification in high 93% yield. In case of enoate 9k, its low solubility in methanol, allowed for efficient isolation of the product by simple filtration. The 4-cyanocinnamate 9i was synthesized by Heck reaction from 4-bromobenzonitrile under low catalyst loading and ligand-free conditions, using Pd(OAc)₂ as catalyst.²² This method is especially efficient for electron deficient aromatic bromides and enoate 9i was obtained in essentially quantitative yield. This result stipulated the choice of 4-bromobenzaldehyde as a precursor for preparation of the CH₂OTBS-substituted enoate 9g, that was finally done in three steps in high overall yield. Thus, the Heck reaction to (E)-methyl 4-formylcinnamate (13) was followed by NaBH₄ reduction of the aldehyde group and silylation of the primary alcohol. Cinnamate 9h with a CH₂CH₂OTBS substituent in 4-position was prepared from the arylacetic acid 14 through reduction with LiAlH₄ and silylation of the resulting primary alcohol. A subsequent halogen-metal exchange on aryl bromide 15 and quenching the anion with DMF furnished aldehyde 16 which was extended via Wittig reaction in good yield. The synthesis of the last derivative, enoate 9h initially was also attempted using Heck reaction, however the corresponding substrate 15 did not react neither under ligand free conditions nor in the presence of PPh₃. Here, formation of palladium black was observed and starting material was recovered unchanged. This is explained by the low reactivity of the aromatic bromide, which is not activated by an electron-withdrawing substituent. In all cases the cinnamates 9 were obtained configurationally pure and no cis-isomers could be detected by NMR in the products.

With several cinnamic esters in hands we next turned to the synthesis of amino esters 8 (Scheme 3). Sharpless asymmetric dihydroxylation (AD)²³ under standard conditions (0.4mol % Os catalyst, 2mol % ligand) was used initially, but this proved problematic providing the diols only in moderate yield (30–68%). After some experimentation, possible reasons for these problems became clear. Due to the electron-deficient nature of the alkenes, especially of the nitro and cyano derivatives, dihydroxylation occurred with reduced rate and saponification of the methyl ester became a significant side reaction. In the case of the nitro substituted diol 17e the corresponding acid was partially isolated and characterized (see the supporting information). The dihydroxylation reactions with the NO₂-, Ph-, and CN-substituted cinnamates were additionally hampered by very limited solubility of the alkenes in the reaction medium. As soon as the ester saponification seemed to be the only problem and otherwise the reaction was clean, attempts were made to improve the yield by addition of KHCO₃ to the reaction mixture, to make it less alkaline, but this was again not sufficiently successful. Finally, the situation was significantly improved when increased catalyst loading was used together with buffering the reaction mixture with KHCO₃. Indeed, in the presence of 1 mol % K₂OsO₂(OH)₄/1.5 mol % chiral ligand, the reaction required somewhat shorter time and the diol 17g was obtained in 93% yield. Although the synthesis of other diols was not attempted under the optimised conditions, we believe that increasing the catalyst loading (e.g. up to 2 mol %) should be the most efficient way to increase the yield in the AD of these relatively unreactive and hardly soluble alkenes. This is in line with the observations given in the literature.²³ For three of the diols (17e, 17g, 17i) the optical purity was assessed by means of HPLC on a chiral column (see the supporting information), and in no case the opposite enantiomer could be detected, which implies at least 98% ee for these compounds.¹⁶ For the other diols the enantiomeric purity was tentatively assessed to be “high” by the observation of no significant side peaks in the ¹H and ¹³C NMR spectra on later steps when diastereomers would appear.

Scheme 3.

Conversion of the cinnamates 9 to the β-tyrosine derivatives 19, respectively; DCE = dichloroethane, DEAD = diethyl azodicarboxylate.

Diols 17 were next transformed to β-azides 18 by means of a Mitsunobu reaction with hydrazoic acid. ²⁴ Initial experiments revealed only moderate conversion of the diols, accompanied by significant nitrogen evolution. This side reaction was caused by hydrazoic acid decomposition of the ternary mixture in the presence of triphenyl phosphine and DEAD. Mixing all three components together caused rapid nitrogen evolution, while the binary mixture of HN₃ and PPh₃ was stable at room temperature. This undesired side reaction was diminished when the DEAD solution was introduced into the reaction as the last component at −25 °C rather than at 0 °C. Thereafter, the mixture was allowed to stand at room temperature until the reaction stopped. Under these conditions, the desired azides were obtained in better yields as pure stereoisomers in accordance with literature data.²⁴ The structure and purity were additionally confirmed by careful comparison of NMR spectra with previously described compounds (see supporting information). In these Mitsunobu reactions, diols 17e (NO₂), 17i (CN), and 17k (Ph) were used as hardly separable mixtures, slightly contaminated with MeSO₂NH₂, resulting from the previous step. We found MeSO₂NH₂ to have no influence on the reaction.

Subsequently, azides 18 were smoothly methylated by trimethyloxonium tetrafluoroborate in the presence of proton sponge to give the α-methoxy propanoates 19. The cyano derivative 19i was hydrolyzed with H₂O₂/K₂CO₃ to provide amide 19j, the precursor for chondramide analogue 2j. The hydrolysis was performed in DMSO, MeOH and DMF media and was found to be the fastest in DMSO (complete conversion was observed already in 1–2 minutes), but in this case the product was isolated only as inseparable mixture with dimethyl sulfone as by-product. Pure amide 19j was obtained in MeOH, but in this case the reaction took much longer time and gave the product in reduced yield most likely due to hydrolysis of the ester.

Azides 19 were conveniently hydrogenated in presence of Pd/C to afford the amino esters 8. Nitro derivative 19i was selectively reduced by treatment with PPh₃ followed by hydrolysis of the iminophosphorane intermediate to afford amine 8e, that was used without further purification.

In several steps amino esters were transformed into chondramide analogs 2b–k (Scheme 4). First, peptide coupling with the known dipeptide acid^25,11c 6 afforded tripeptides 20 in good yields. Subsequent saponification of the methyl ester, either with Me₃SnOH (20b–d) or simply with NaOH [20e (NO₂), 20i (CN), 20j (CONH₂), 20k (Ph)] occurred in quantitative yields and without epimerization. But saponification of the CH₂OTBS- and (CH₂)₂OTBS-substituted tripeptide esters 20g and 20h was accompanied by partial deprotection of the alcohol functions. Still, pure acids 21g and 21h were successfully isolated by chromatography in reduced, but good yields (~80%). Thus, there was no need to call upon the milder Me₃SnOH procedure. On the next stage, acids 21b–d were initially esterified with hydroxy ester 7-epi-7 under Yamaguchi conditions to give the seco-compounds 22b–d in reasonable yields. However, later we found it more convenient to perform the esterification reaction under Mitsunobu conditions since the synthesis of hydroxy ester 7 is shorter than the one for 7-epi-7.¹⁵ Therefore, the acids 21 were esterified with alcohol¹⁵ 7 under Mitsunobu conditions in quite good yields. Here, the correct use of either DEAD or DIAD reagents in each case is important for successful separation of the esterification product. Otherwise, the corresponding hydrazodicarboxylate by-product would be inseparable (or hardly separable) from the product. Unfortunately, the acid 21j (CONH₂) could not be esterified under Mitsunobu conditions. Other methods (Yamaguchi esterification, DCC/DMAP conditions) were unsuccessful either. Eventually, the corresponding chon A derivative was prepared by nitrile hydrolysis on the macrocycle 2j (vide infra). For macrolactam formation, the NBoc protected tert-butyl esters 22 were subjected to double deprotection under acidic conditions. Initially, the reaction was carried out in a mixture of TFA/CH₂Cl₂ (1:30, vol/vol) as previously described.¹⁶ The progress of the deprotection was easily monitored by TLC and NMR and under these conditions the conversion was not complete even after 2 d. In particular, the carboxylic ester cleavage is quite slow. However with more concentrated acid (1:10, vol/vol) complete and clean deprotection was observed for each compound after about 20 h. Interestingly, under these reaction conditions the TBS groups of 22g and 22h were quantitatively transformed to the corresponding TFA esters, as suggested by NMR of the crude products. In each case the crude amino acids were subjected to macrolactamization in the presence of TBTU, which afforded chondramide A analogs 2e (NO₂), 2i (CN), and 2k (Ph) directly. In order to obtain the analogues 2g (CH₂OH) and 2h [(CH₂)₂OH] the crude cyclization products were stirred in aqueous K₂CO₃/MeOH for 5–10 minutes before purification. This resulted in complete hydrolysis of the corresponding TFA esters. Unfortunately, the presence of the benzylic trifluoroacetate negatively affected cyclization of the amino acid generated from 22g (CH₂OTBS), yielding after short hydrolysis with aqueous K₂CO₃/MeOH two products as judged by TLC. The more polar substance turned out to be the desired chondramide A analogue 2g. The second substance was shown to be the corresponding benzotriazolyl ether that was characterized by ¹H NMR and HRMS analysis (see supporting information for details). Thus, for this analogue it can be concluded that the TBS group is not ideal and revision of the synthetic scheme would be required in this case for an optimized synthesis. But for the other cases the described synthetic scheme is considered reliable. With the NO₂-substituted analogue 2e we attempted reduction of the nitro group to obtain the NH₂ derivative 2f. This was tried with catalytic hydrogenation and sodium dithionite. Both methods afforded the desired amino chondramide 2f in moderate yield. Finally, the CONH₂-substituted analogue 2j was prepared by hydrolysis of the cyano group of 2i with H₂O₂/K₂CO₃/DMSO under mild conditions in good unoptimised yield.

Scheme 4.

Synthesis of chondramide A analogues 2b–j. Esters 22b–d were also prepared by Yamaguchi esterification using hydroxy ester 7-epi-7; TBTU = 2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethylaminium tetrafluoroborate, HOBt = 1-hydroxy-benzotriazole, DIAD = diisopropyl azodicarboxylate. [a] for compound 2, R = CH₂OH and (CH₂)₂OH, respectively.

Biology

The biological activity of the chondramide A (chon A) analogues 2 was evaluated in a cytotoxicity assay on human primary foreskin fibroblast (HFF) cells. These data are included in Table 1. We also included the natural products jasplakinolide and other chondramides in this assay. As can be seen, several of the chon A analogues are slightly more active than the natural product itself. Three of the analogues [2b (Me), 2d (F), 2i (CN)] are even comparable to jasplakinolide (1) itself (EC₅₀ in the range of 30 nM). These are characterized by small substituents. Other analogues show intermediate activity, being about half as active as jasplakinolide (EC₅₀ in the range of 60 nM). These analogues include 2c (OMe), 2e (NO₂), 2f (NH₂), 2g (CH₂OH), and 2h ((CH₂)₂OH). Somewhat surprising is the activity of analogue 2k (Ph) (EC₅₀ = 213 nM). With this large substituent one would not expect that the analogue would fit in the binding pocket.¹² Although the amide substituent in 2j is not very large, this compound is much less active. A similar trend in activity can be seen on the mouse fibroblast cell line L-929, which is more sensitive to these cyclodepsipeptides.

Table 1.

Cytotoxicity of jasplakinolide (1) and chondramide A derivatives 2.

entry	compound	EC50 (HFF) [μM]	IC50 (L929) [μM]
1	jasplakinolide_np (1)	0.028±0.004
2	chon A_np (2a)	0.118±0.023	0.056±0.013
3	chon B_np (3)	0.044±0.005
4	chon C_np (4)	0.036±0.007
5	chon A_synth (2a)	0.109±0.007
6	chon A_Me (2b)	0.027±0.002	0.0079±0.0013
7	chon A_OMe (2c)	0.066±0.008	0.0083±0.0021
8	chon A_F (2d)	0.029±0.002	0.0080±0.0013
9	chon A_NO2 (2e)	0.040±0.006	0.017±0.005
10	chon A_NH2 (2f)	0.057±0.001	0.031±0.002
11	chon A_CH2OH (2g)	0.052±0.010	0.035±0.002
12	chon A_(CH2)2OH (2h)	0.059±0.010	0.033±0.004
13	chon A_CN (2i)	0.027±0.001	0.011±0.001
14	chon A_CONH2 (2j)	1.424±0.017	0.223±0.042
15	chon A_Ph (2k)	0.213±0.016	0.025±0.002

np = natural product (isolated), synth = synthetic compound.

In general, it can be concluded that the aryl ring on the tyrosine derivative accepts a range of substituents. One explanation could be that the 4-position of the aryl ring protrudes out of the binding pocket. Thus, some of the prepared derivatives might be used for the preparation of labeled derivatives. The recent finding that jasplakinolide V with a catechol in place of the phenol substituent is also quite active shows this region to be less sensitive to structural modifications than the other parts.^8b

Conclusion

Based on our previous synthesis of chondramide A, a range of analogues was prepared that feature modifications in the 4-position of the β-tyrosine derivative. The required building blocks, 3-aryl-3-amino-2-methoxypropanoates 8 were prepared from cinnamates 9 by asymmetric dihydroxylation, regioselective Mitsunobu substitution with hydrazoic acid, O-methylation and reduction of the azide group. Subsequently, amino esters 8 were condensed with dipeptide acid 6. After hydrolysis of the methyl ester function, esterification of the tripeptide acids 21 with the 7-hydroxy ester 7 or 7-epi-7, respectively furnished the seco-compounds 22. Deprotection of both the amino and the ester group was followed by macrolactam formation to give analogues 2.

In cell growth assays several of the derivatives surpassed natural chondramide A in its biological activity. Among the derivatives only the amide derivative 2j was not very active. Otherwise, it appears that the aryl ring of the β-tyrosine tolerates a broad range of substituents in the 4-position. These derivatives should be useful for probing a number of cellular questions in different systems that rely on actin filaments for important aspects of biology.

Experimental Section

Only part of the experimental material is given here. For further details and for compound characterization see Supporting Information. Procedures are given for the sequence leading to chondramide A analogue 2e (NO₂) and the reduction of 2e to analogue 2f (NH₂).

(E)-Methyl 4-nitrocinnamate (9e) (by Knoevenagel reaction)

A solution of nitrobenzaldehyde (1.51 g, 0.01 mol), malonic acid (1.14 g, 1.1 equiv), and pyridine (0.25 mL, 0.31 equiv) in ethanol (2 mL) was heated at reflux. Already after 10–15 min a white solid began to precipitate. After 1.5 h, when TLC (EtOAc/petroleum ether, 1:2) indicated complete consumption of the aldehyde, the reaction mixture was cooled and acidified with diluted aqueous HCl. The precipitate was filtered, washed with water and dried in vacuo to afford crude but pure (2E)-3-(4-nitrophenyl)acrylic acid in quantitative yield (note: the free acid is soluble in acetone, DMSO, but hardly soluble in CH₂Cl₂, EtOAc). It was taken up in methanol (20 mL), thionyl chloride (1.3 equiv) was carefully added and the resulting suspension was heated at reflux for 1 h. At this time TLC (EtOAc/petroleum ether, 1:2) indicated complete consumption of the acid and the volume of precipitates visually increased. The product was obtained by crystallization directly from the reaction mixture as follows. The reaction mixture was concentrated in vacuo till a thick suspension resulted, that was filtered on a glass frit and the precipitate was washed with a small amount of cold methanol to afford 1.52 g (93%) of pure methyl (2E)-3-(4-nitrophenyl)acrylate (9e) as a slightly yellow solid. Analytical data were in agreement with literature data.²⁶

(2S,3R)-Methyl 3-(4-cyanophenyl)-2,3-dihydroxypropanoate (17e)

A mixture of K₃Fe(CN)₆ (7.22 g, 22.0 mmol, 3 equiv), K₂CO₃ (3.04 g, 22.0 mmol, 3 equiv), MeSO₂ NH₂ (0.70 g, 7.37 mmol, 1 equiv), K₂OsO₂(OH)₄ (10 mg, 0.027 mmol, 0.0037 equiv), and the ligand (DHQD)₂PHAL (57 mg, 0.073 mmol, 0.01 equiv) was stirred in a mixture of water (35 mL) and tBuOH (35 mL) until dissolved and then the solution was cooled to 0 °C in an ice bath. At this point cinnamate 9e (1.52 g, 7.34 mmol) was added and the reaction mixture was allowed to reach room temperature slowly while being stirred overnight, at which time a yellow suspension was formed and complete or almost complete conversion was observed according to TLC (petroleum ether/EtOAc, 1:1). Then solid Na₂SO₃ (9.2 g, 73.4 mmol, 10 equiv) was added and the mixture was stirred for several min. The suspension was filtered, and the filter cake was washed with EtOAc. The filtrate was transferred to a separatory funnel, the organic phase was separated, and the water phase was extracted twice with EtOAc. The combined organic extracts were washed with saturated NaCl solution, dried with Na₂SO₄, filtered, and evaporated. The residue was purified by flash chromatography (CH₂Cl₂/MeOH, 95:5) to afford pure diol 17e (0.698 g, 39%) as a colorless solid. Sometimes it was difficult to purify the diol from CH₃SO₃NH₂ (as a contaminant in different compounds comes at δ = 3.08–3.10 (s, 3H), 4.67–4.79 (bs, 2H)), but the sulfone amide impurity did not influence the next Mitsunobu azidation as was established later. Also, a better procedure but for another substrate (17g) utilizing 1% of the catalyst is described in the Supporting information. ¹H NMR (400 MHz, CDCl₃): δ = 2.81 (bs, 1H, OH), 3.13 (bs, 1H, OH), 3.86 (s, 3H, OCH₃), 4.39 (bs, 1H, 2-H), 5.14 (bs, 1H, 3-H), 7.59 (d, J = 8.4 Hz, 2H, Ar), 8.23 (d, J = 8.4 Hz, 2H, Ar).

Preparation of hydrazoic acid solution: Caution

Hydrazoic acid is a highly volatile, toxic and explosive liquid in individual state. However, in solution it is stable and safe. In this study solutions up to 5M were used. Sodium azide (3.0 g, 46 mmol) was mixed with water (1.5 mL) and toluene (10 mL). The suspension was cooled to near 0 °C and concentrated H₂SO₄ (~2 g, ~1.09 mL, 20 mmol) was carefully added while cooling and shaking the round bottom flask (stirring with magnetic stirring bar is not sufficient). Crystals were kneaded with a spatula shortly after the addition of the acid. Then the mixture was filtered under positive pressure and dried with Na₂SO₄ (there was no water phase remained, and no need to separate it from toluene solution). To rapidly estimate the resulting concentration of hydrazoic acid, a known amount of NaOH was dissolved in a small amount of water, phenolphthalein was added and the pink solution was titrated with the hydrazoic acid solution from an analytical pipette (the concentration was found to be 3.3M against theoretical 4.0M).

(2S,3S)-Methyl 3-azido-2-hydroxy-3-(4-nitrophenyl)propanoate (18e)

To a stirred solution of diol 17e (0.228 g, 0.946 mmol), triphenylphosphine (0.297 g, 1.14 mmol, 1.2 equiv), hydrazoic acid (0.86 mL, 3.3M in toluene, 3 equiv) in THF (2.0 mL) at −25 °C was added DEAD (0.56 mL, 0.535 g, 1.23 mmol, 40% wt. solution in toluene, 1.3 equiv), then the cooling bath was removed (slight evolution of N₂ was observed) and the resulting mixture was stirred overnight at ambient temperature (TLC control: petroleum ether/EtOAc, 1:1; NMR control: a sample portion was taken from the reaction mixture, evaporated and directly analyzed by NMR). Then the reaction mixture was concentrated in vacuo. The residue was purified by flash chromatography (petroleum ether/EtOAc, 4:1 to 2:1) to yield 3-azido propanoate 18e (0.145 g, 58%) as a slightly orange oil which solidified into a waxy solid upon standing. R_f (petroleum ether/EtOAc, 1:1) 0.55; [α]²⁰_D = +101.1 (c 1.00, CH₂Cl₂); ¹H NMR (400 MHz, CDCl₃): δ = 3.11 (bs, 1H, OH), 3.73 (s, 3H, OCH₃), 4.58 (d, 1H, 2-H), 5.08 (d, J = 3.8 Hz, 1H, 3-H), 7.53 (d, J = 8.7 Hz, 2H, Ar), 8.23 (d, J = 8.7 Hz, 2H, Ar); ¹³C NMR (100 MHz, CDCl₃): δ = 53.1 (OCH₃), 66.4 (C-3), 73.7 (C-2), 123.7 (C_ar), 128.7 (C_ar), 141.8 (C_ar), 148.1 (C_ar), 171.3 (CO₂CH₃); HMRS (ESI): m/z: calcd for C₁₀H₁₀N₄O₅ [M+Na]⁺: 289.05434; found: 289.05431. Note. The product 18e was slightly contaminated with diethyl hydrazodicarboxylate (~3–5 mol %), having signals δ = 1.28±0.01 (t, 6H), 4.22±0.01 (q, 4H). This impurity did not cause problems in the subsequent steps.

(2S,3S)-Methyl 3-azido-2-methoxy-3-(4-nitrophenyl)propanoate (19e)

To a solution of α-hydroxy ester 18e (0.495 g, 1.86 mmol) in dry 1,2-dichloroethane (1.9 mL) was added trimethyloxonium tetrafluoroborate (0.495 g, 3.35 mmol, 1.8 equiv) and proton sponge (0.876 g, 4.09 mmol, 2.2 equiv). The flask was covered with alumina foil. After stirring the suspension at 40 °C overnight, a small probe was taken from the reaction mixture and quenched with EtOAc/HCl_aq for TLC (petroleum ether/EtOAc, 1:1), that indicated full conversion. The reaction mixture was cooled, treated with EtOAc/H₂O, and acidified with 1N HCl to pH 2–3. The precipitate was filtered off and the filtrate was separated. The aqueous phase was extracted once with EtOAc and the combined organic extracts were washed with water, and saturated NaCl solution, dried with Na₂SO₄, filtered, and concentrated in vacuo. The residue was chromatographed (petroleum ether/EtOAc, 3:1 to 2:1) to yield methyl ether 19e (0.48 g, 92%) as a slightly orange oil. R_f (petroleum ether/EtOAc, 2:1) 0.54; [α]²⁰_D = +49.5 (c 1.00, CH₂Cl₂); ¹H NMR (400 MHz, CDCl₃): δ = 3.38 (s, 3H, OCH₃), 3.75 (s, 3H, CO₂CH₃), 3.99 (d, J = 6.1 Hz, 1H, 2-H), 4.89 (d, J = 6.1 Hz, 1H, 3-H), 7.55 (d, J = 8.7 Hz, 2H, Ar), 8.22 (d, J = 8.7 Hz, 2H, Ar); ¹³C NMR (100 MHz, CDCl₃): δ = 52.5 (OCH₃), 59.3 (OCH₃), 65.0 (C-3), 83.0 (C-2), 123.8 (C_ar), 129.0 (C_ar), 142.4 (C_ar), 148.1 (C_ar), 169.6 (CO₂CH₃); HMRS (ESI): m/z: calcd for C₁₁H₁₂N₄O₅ [M+Na]⁺: 303.06999; found: 303.07002.

Reduction of azide 19e and coupling of amine 8e with acid 6 to tripeptide 20e (NO₂)

A solution of azide 19e (89.6 mg, 0.32 mmol) and PPh₃ (92.2 mg, 0.352 mmol, 1.1 equiv) in of THF (1 mL) was stirred at 40–50 °C for 1 h for clean and complete conversion to the corresponding iminophosphorane (TLC control: CH₂Cl₂/MeOH/NH₃, 10:1:0.1, R_f 0.5 for the iminophosphorane). Selected ¹H NMR (400 MHz, CDCl₃) data for the iminophosphorane: δ = 7.91 (d, 2H), 3.67 (s, 3H), 3.13 (s, 3H). Then water (0.1 mL) was added and the mixture was further stirred at 40–50 °C for ~8 h. Because the R_f values of the iminophosphorane, Ph₃P=O and the resulting amine were all the same, the reaction progress was conveniently monitored by analyzing small evaporated probes taken from the reaction mixture by NMR. Selected ¹H NMR (400 MHz, CDCl₃) data for the amine: δ = 8.15 (d, 2H), 3.65 (s, 3H), 3.39 (s, 3H). When appropriately clean and high (~86%) conversion was achieved, the mixture was evaporated to yield 0.166 g of a sticky oil, containing ~36% w/w amine 8e (assuming the conversion was 80% as the lowest). Then, to a solution of this crude mixture (108 mg), containing amine 8e (approx. 38.9 mg, 0.153 mmol, 1.24 equiv of the amine) in DMF (2.3 mL) were added acid 6 (47.7 mg, 0.123 mmol), HOBt (24.9 mg, 0.184 mmol, 1.5 equiv), iPr₂NEt (0.064 mL, 0.369 mmol, 3 equiv). At −10 °C TBTU (59 mg, 0.184 mmol, 1.5 equiv) was added and the reaction was stirred for 5–6 h at room temperature. The mixture was diluted with water (5 mL) and extracted with ethyl acetate (3 × 8 mL). The combined organic layers were washed with 1N NaHSO₄ solution (5 mL), saturated NaHCO₃ solution (5 mL), saturated NaCl solution (5 mL), dried with Na₂SO₄, filtered, and concentrated in vacuo. Purification of the residue by flash chromatography (petroleum ether/EtOAc, 2:1) gave tripeptide 20e (58.7 mg, 76%) as a white foam. R_f (petroleum ether/EtOAc, 2:1) 0.48. [α]²⁰_D = +4.0 (c 1.00, CH₂Cl₂); ¹H NMR (400 MHz, CDCl₃): δ = 0.97 (d, J = 6.9 Hz, 3H, Ala CH₃), 1.40 (s, 9H, tBu), 2.98 (s, 3H, NCH₃), 3.17 (dd, J = 15.5, 9.4 Hz, 1H, CH₂), 3.38 (s, 3H, OCH₃), 3.38 (m, 1H, CH₂), 3.61 (s, 3H, CO₂CH₃), 4.07 (d, J = 5.1 Hz, 1H, CHOCH₃), 4.48–4.54 (m, 1H, Ala CH), 5.33 (d, J = 7.4 Hz, 1H, Ala NH), 5.44 (dd, J = 8.1, 5.1 Hz, 1H, β-Tyr CH), 5.53 (dd, J = 8.9, 7.4 Hz, 1H, Trp CH), 6.89 (s, 1H, Trp H_Ar), 7.08 (ddd, J = 7.9, 7.1, 0.8 Hz, 1H, Trp H_Ar), 7.16 (ddd, J = 7.9, 7.1, 0.8 Hz, 1H, Trp H_Ar), 7.21 (d, J = 8.1 Hz, 1H, β-Tyr NH), 7.31 (d, J = 7.9 Hz, 1H, Trp H_Ar), 7.34 (d, J = 8.3 Hz, 2H, Ar), 7.57 (d, J = 7.9 Hz, 1H, Trp H_Ar), 8.06 (d, J = 8.3 Hz, 2H, Ar), 8.21 (s, 1H, Trp NH); ¹³C NMR (100 MHz, CDCl₃): δ = 17.8 (Ala CH₃), 23.5 (CH₂), 28.3 (C(CH₃)₃), 30.9 (NCH₃), 46.7 (Ala CH), 52.1 (OCH₃), 53.8 (β-Tyr CH), 56.8 (Trp CH), 59.3 (OCH₃), 79.7 (C(CH₃)₃), 81.8 (CHOCH₃), 110.5 (quat. Trp), 111.2, 118.5, 119.5, 122.2 (2C, Trp), 123.4 (2C, Ar), 127.1 (quat. Trp), 128.7 (2C, Ar), 136.1 (quat. Trp), 144.3 (quat. Ar), 147.5 (quat. Ar), 155.2 (Boc), 169.5, 169.6, 174.5; HMRS (ESI): m/z: calcd for C₃₁H₃₉N₅O₉ [ M+Na]⁺: 648.26400; found: 648.26502.

Tripeptide acid 21e (NO₂)

To a solution of methyl ester 20e (55.3 mg, 0.0884 mmol) in THF (0.4 mL) were added water (0.6 mL), methanol (0.3 mL) and NaOH (7.5 mg, 0.188 mmol, 2.1 equiv). The initial biphasic mixture became homogeneous with progressing saponification. After being stirred for 1 h at room temperature until complete conversion (controlled by TLC), the mixture was diluted with water (5 mL) and ethyl acetate (8 mL). It was carefully acidified with 1M NaHSO₄ to pH ~ 2 before the layers were separated and the aqueous phase extracted once with ethyl acetate (8 mL). The combined organic layers were washed with water, saturated NaCl solution, dried with Na₂SO₄, filtered, and concentrated in vacuo to afford the crude acid 21e as a colorless foam. R_f (EtOAc/AcOH, 100:1) 0.4; ¹H NMR (400 MHz, CDCl₃): δ = 0.83 (d, J = 6.9 Hz, 3H, Ala CH₃), 1.39 (s, 9H, t-Bu), 2.93 (s, 3H, NCH₃), 3.18 (dd, J = 15.3, 9.9 Hz, 1H, CH₂), 3.28–3.34 (m, 1H, CH₂), 3.34 (s, 3H, OCH₃), 3.98 (d, J = 6.3 Hz, 1H, CHOCH₃), 4.42–4.49 (m, 1H, Ala CH), 5.40 (dd, J = 7.6, 7.1 Hz, 1H, β-Tyr CH), 5.46 (d, J = 7.4 Hz, 1H, Ala NH), 5.59 (dd, J = 9.9, 7.4 Hz, 1H, Trp CH), 6.88 (s, 1H, Trp H_Ar), 7.05 (app t, J = 7.0 Hz, 1H, Trp H_Ar), 7.13 (app t, J = 7.0 Hz, 1H, Trp H_Ar), 7.21 (d, J = 8.6 Hz, 1H, β-Tyr NH), 7.28 (d, J = 7.9 Hz, 1H, Trp H_Ar), 7.41 (d, J = 8.4 Hz, 2H, Ar), 7.53 (d, J = 7.9 Hz, 1H, Trp H_Ar), 8.04 (d, J = 8.4 Hz, 2H, Ar), 8.36 (s, 1H, Trp NH); ¹³C NMR (100 MHz, CDCl₃): δ = 17.0 (Ala CH₃), 23.2 (CH₂), 28.3 (C(CH₃)₃), 30.6 (NCH₃), 46.6 (Ala CH), 54.0 (β-Tyr CH), 56.7 (Trp CH), 58.8 (OCH₃), 80.8 (C(CH₃)₃), 82.5 (CHOCH₃), 110.2 (quat. Trp), 111.2, 118.4, 119.5, 122.1, 122.3, 123.5 (2C, Ar), 127.1 (quat. Trp), 128.4 (2C, Ar), 136.1 (quat. Trp), 145.2 (quat. Ar), 147.4 (quat. Ar), 156.3 (Boc), 169.5, 170.7, 174.8; HMRS (ESI): m/z: calcd for C₃₀H₃₇N₅O₉ [M+Na]⁺: 634.24835; found: 634.24819.

Depsipeptide 22e (NO₂)

The crude acid 21e (52 mg, 0.0851 mmol) and alcohol 7 (32.8 mg, 0.128 mmol, 1.5 equiv) were dissolved in THF (1 mL) and Ph₃P (40 mg, 0.152 mmol, 1.8 equiv) was added at 0 °C. This was followed by the dropwise addition of DIAD (0.030 mL, 0.152 mmol, 1.8 equiv). The cooling bath was removed and the mixture stirred overnight at room temperature. The reaction mixture was concentrated in vacuo and the residue purified by flash chromatography (petroleum ether/EtOAc, 2:1 to 1:1) to give ester 22e (62 mg, 86%) as a colorless foam. R_f (petroleum ether/EtOAc, 1:1) 0.30; [α]²⁰_D = −4.0 (c 1.00, CH₂Cl₂); ¹H NMR (400 MHz, CDCl₃): δ = 0.77 (d, J = 6.6 Hz, 3H, CH₃), 0.91 (d, J = 6.1 Hz, 3H, CH₃), 0.96 (d, J = 6.8 Hz, 3H, Ala CH₃), 1.01 (d, J = 6.8 Hz, 3H, CH₃), 1.40 (s, 9H, t-Bu), 1.41 (s, 9H, t-Bu), 1.56 (s, 3H, CH₃), 1.92 (dd, J = 13.9, 7.6 Hz, 1H, CH₂), 2.33 (dd, J = 13.9, 6.8 Hz, 1H, CH₂), 2.40–2.51 (m, 2H, 2 CH), 2.97 (s, 3H, NCH₃), 3.17 (dd, J = 15.4, 9.6 Hz, 1H, CH₂), 3.37 (dd, J = 15.4, 7.1 Hz, 1H, CH₂), 3.40 (s, 3H, OCH₃), 4.06 (d, J = 4.5 Hz, 1H, CHOCH₃), 4.46–4.61 (m, 2H, Ala CH, CO₂CH), 4.84 (d, J = 9.9 Hz, 1H, =CH), 5.34 (d, J = 7.3 Hz, 1H, Ala NH), 5.43 (dd, J = 8.3, 4.5 Hz, 1H, β-Tyr CH), 5.52 (dd, J = 9.6, 7.1 Hz, 1H, Trp CH), 6.88 (s, 1H, Trp H_Ar), 7.07 (ddd, J = 7.8, 7.1, 0.8 Hz, 1H, Trp H_Ar), 7.15 (ddd, J = 8.1, 7.1, 0.8 Hz, 1H, Trp H_Ar), 7.21 (d, J = 8.3 Hz, 1H, β-Tyr NH), 7.30 (d, J = 8.1 Hz, 1H, Trp H_Ar), 7.39 (d, J = 8.6 Hz, 2H, Ar), 7.56 (d, J = 7.8 Hz, 1H, Trp H_Ar), 8.04 (d, J = 8.6 Hz, 2H, Ar), 8.25 (s, 1H, Trp NH); ¹³C NMR (100 MHz, CDCl₃): δ = 16.4, 16.6, 17.1, 17.6, 17.9, 23.5 (CH₂), 28.0 (C(CH₃)₃), 28.3 (C(CH₃)₃), 30.9 (NMe), 37.5, 38.6, 43.4 (CH₂), 46.7 (Ala CH), 53.8 (β-Tyr CH), 56.8 (Trp CH), 59.3 (OCH₃), 76.4 (CO₂CH), 79.6 (C(CH₃)₃), 79.9 (C(CH₃)₃), 81.4 (CHOCH₃), 110.5 (quat. Trp), 111.1, 118.4, 119.5, 122.1, 122.2, 123.3 (2C, Ar), 127.1 (quat. Trp), 127.5 (=CH), 129.1 (2C, Ar), 134.0 (=C<), 136.1 (quat. Trp), 144.3, 147.4, 155.2 (Boc), 168.6, 169.5, 174.5, 175.7; HMRS (ESI): m/z: calcd for C₄₅H₆₃N₅O₁₁ [M+Na]⁺: 872.44163; found: 872.44212.

Chondramide 2e (NO₂)

To a stirred solution of compound 22e (62.7 mg, 0.0738 mmol) in CH₂Cl₂ (2.2 mL) was added TFA (0.22 mL, 0.34 g, 2.98 mmol) at 0 °C. The reaction mixture was allowed to warm to room temperature and after stirring for 22 h, the solvent was removed in vacuo (TLC control: CH₂Cl₂/MeOH/NH₃, 10:1:0.1). For azeotropic removal of TFA the residue was taken up in toluene (3 × 0.5 mL) and concentrated in vacuo each time. The crude product was dissolved in DMF (30 mL) and iPr₂NEt (64 μL, 47.7 mg, 0.369 mmol, 5 equiv), HOBt (19.9 mg, 0.147 mmol, 2 equiv) and TBTU (47.4 mg, 0.147 mmol, 2 equiv) were added. The solution was stirred at room temperature for 20 h and then diluted with water (20 mL) and EtOAc (20 mL). The aqueous layer was extracted with EtOAc (3 × 20 mL) and the combined organic layers were washed with 5% aqueous KHSO₄ solution (20 mL), water (20 mL), saturated NaHCO₃ solution (20 mL), water (2 × 20 mL) and saturated NaCl solution (20 mL). The combined organic extracts were dried over Na₂SO₄, filtered, and concentrated in vacuo. The crude product was purified by flash chromatography (petroleum ether/EtOAc, 1:3 to 0:1) to give depsipeptide 2e (27.3 mg, 48%) as a colorless foam. R_f (EtOAc) 0.41; [α]²⁰_D = +18.5 (c 1.00, CH₂Cl₂); ¹H NMR (400 MHz, CDCl₃): δ = 0.84 (d, J = 6.6 Hz, 3H, CH₃), 0.85 (d, J = 6.1 Hz, 3H, CH₃), 1.07 (d, J = 6.8 Hz, 3H, CH₃), 1.17 (d, J = 6.8 Hz, 3H, CH₃), 1.64 (s, 3H, CH₃), 1.88 (d, J = 13.1 Hz, 1H, CH₂), 2.36–2.51 (m, 3H, 2 CH, CH₂), 2.95 (s, 3H, NCH₃), 3.17 (dd, J = 15.2, 8.6 Hz, 1H, CH₂), 3.22 (s, 3H, OCH₃), 3.31 (dd, J = 15.2, 7.8 Hz, 1H, CH₂), 3.75 (d, J = 8.1 Hz, 1H, CHOCH₃), 4.77–4.83 (m, 2H, Ala CH, CO₂CH), 4.90 (d, J = 9.1 Hz, 1H, =CH), 5.30 (t, J = 8.3 Hz, 1H, β-Tyr CH), 5.60 (t, J = 8.1 Hz, 1H, Trp CH), 6.50 (d, J = 7.3 Hz, 1H, Ala NH), 6.82 (s, 1H, Trp H_Ar), 7.06 (d, J = 8.8 Hz, 1H, β-Tyr NH), 7.10 (ddd, J = 8.1, 7.1, 0.8 Hz, 1H, Trp H_Ar), 7.17 (ddd, J = 8.1, 7.1, 0.8 Hz, 1H, Trp H_Ar), 7.24 (d, J = 8.6 Hz, 2H, Ar), 7.32 (d, J = 8.1 Hz, 1H, Trp H_Ar), 7.59 (d, J = 8.1 Hz, 1H, Trp H_Ar), 8.06 (d, J = 8.6 Hz, 2H, Ar), 8.11 (s, 1H, Trp NH); ¹³C NMR (100 MHz, CDCl₃): δ = 15.8, 16.6, 17.3, 18.6, 20.0, 23.6 (CH₂), 30.2 (NCH₃), 37.0, 40.2, 44.1 (CH₂), 45.2 (Ala CH), 54.4 (β-Tyr CH), 56.0 (Trp CH), 58.2 (OCH₃), 76.8 (CO₂CH), 81.9 (CHOCH₃), 110.2 (quat. Trp), 111.2, 118.5, 119.6, 122.2, 122.3, 123.4, 127.1 (quat. Trp), 127.8 (=CH), 127.9, 134.3 (=C<), 136.1 (quat. Trp), 144.5, 147.3, 169.77, 169.82, 174.1, 174.5; HMRS (ESI): m/z:calcd for C₃₆H₄₅N₅O₈ [M+Na]⁺: 698.31603; found: 698.31672.

Chondramide 2f (NH₂)

By catalytic hydrogenation

A solution of chondramide 2e (6.1 mg, 9.03 μmol) in methanol (0.5 mL) was hydrogenated overnight in a round bottom flask connected to a hydrogen filled balloon, using 10% Pd on carbon (TLC control: CH₂Cl₂/MeOH, 10:1). Upon complete conversion, the solvent was evaporated and the residue purified by flash chromatography (CH₂Cl₂/MeOH, 30:1 to 20:1) to afford chondramide 2f (2.7 mg, 47%) as white foam. A mixed fraction, containing 2f and another unknown compound as major component was also isolated. Complete conversion was observed, but the reaction was not very clean, likely because of formation of RNO and/or RNHOH, as suggested by LC/MS examination of the mixed fraction. No evidence for competitive hydrogenation of the double bond was found (by ¹H NMR).

By reduction with sodium dithionite

Alternatively, reduction of chondramide 2e (5.8 mg, 8.58 μmol) was carried out in THF (0.5 mL), H₂ O (0.5 mL) in the presence of Na₂S₂O₄ (20 mg, 0.115 mmol) overnight at room temperature. Conversion was high but not complete. The reaction mixture was diluted with saturated NaCl solution and EtOAc with addition of some saturated NaHCO₃ solution. The water phase was extracted once more with EtOAc. The combined organic extracts were dried with Na₂SO₄, filtered, and evaporated. The desired NH₂-chondramide 2f (2 mg, 36%) was isolated successfully by flash chromatography (CH₂Cl₂/MeOH 30:1 to 20:1) as a white foam. R_f (CH₂Cl₂/MeOH 20:1) 0.26. [α]²⁰_D = +24.9 (c 0.35, CH₂Cl₂); ¹H NMR (400 MHz, CDCl₃): δ = 0.80 (d, J = 6.3 Hz, 3H, CH₃), 0.85 (d, J = 6.8 Hz, 3H, CH₃), 1.03 (d, J = 6.8 Hz, 3H, Ala CH₃), 1.15 (d, J = 7.1 Hz, 3H, CH₃), 1.63 (s, 3H, CH₃), 1.80 (d, J = 13.9 Hz, 1H, CH₂), 2.33–2.40 (m, 2H, 2 CH), 2.53 (d, J = 10.4, 13.6 Hz, 1H, CH₂), 2.94 (s, 3H, NCH₃), 3.14 (dd, J = 15.5, 9.2 Hz, 1H, CH₂), 3.23 (s, 3H, OCH₃), 3.27 (dd, J = 15.5, 6.8 Hz, 1H, CH₂), 3.68 (bs, 2H, NH₂), 3.75 (d, J = 7.3 Hz, 1H, CHOCH₃), 4.77–4.84 (m, 2H, Ala CH, 7-H), 4.90 (d, J = 8.8 Hz, 1H, =CH), 5.25 (dd, J = 7.3, 9.1 Hz, 1H, β-Tyr CH), 5.60 (dd, J = 7.1, 9.1 Hz, 1H, Trp CH), 6.58 (d, J = 7.3 Hz, 1H, Ala NH), 6.58 (d, J = 8.3 Hz, 2H, Ar), 6.81 (d, J = 2.2 Hz, 1H, Trp H_Ar), 6.92–6.96 (m, 3H, β-Tyr NH, Ar), 7.11 (ddd, J = 8.1, 7.8, 1.0 Hz, 1H, Trp H_Ar), 7.17 (ddd, J = 8.1, 7.8, 1.0 Hz, 1H, Trp H_Ar), 7.30 (d, J = 7.8 Hz, 1H, Trp H_Ar), 7.90 (d, J = 7.8 Hz, 1H, Trp H_Ar), 7.87 (s, 1H, Trp NH); ¹³C NMR (100 MHz, CDCl₃): δ = 15.3, 16.6, 17.6, 18.6, 20.5, 23.3 (CH₂), 30.1 (NCH₃), 37.1, 40.3, 43.8 (CH₂), 45.3 (Ala CH), 53.9 (β-Tyr CH), 55.7 (Trp CH), 58.0 (OCH₃), 77.2 (CO₂CH), 82.6 (CHOCH₃), 110.7 (quat. Trp), 111.0, 115.0 (2C, Ar), 118.6, 119.4, 122.05, 122.11, 127.2 (quat. Trp), 127.7 (quat. Ar), 127.9 (2C, Ar), 128.4 (=CH), 134.1 (=C<), 136.1 (quat. Trp), 145.9, 169.6, 170.6, 174.0, 174.5; HMRS (ESI): m/z: calcd for C₃₆H₄₇N₅O₆ [M+Na]⁺: 668.34186; found: 668.34224.

Cytotoxicity Assay and EC₅₀ Determination

Human foreskin fibroblast (HFF) cells grown in Dulbecco’s modified Eagle’s medium with 10% fetal bovine serum and 10 μg mL⁻¹ gentamicin were seeded at 2,000 cells/well in 96-well microtiter plates. After 2 h of incubation at 37 °C in a humidified atmosphere with 5% CO₂, Jasplakinolide and chon A analogues were added to reach the final concentrations of 0.0049–5 μM (in 2-fold increments). The number of viable cells was determined 72 h later using the CellTiter 96 AQ_ueous One Solution Cell Proliferation Assay (Promega). Absorbance at 570 λ (A₅₇₀) measured was normalized to percentage of control (untreated) and plotted against logarithmic conversions of compound concentrations using Prism 5.0 (GraphPad Software). The log (inhibitor) vs. response curve with variable slope was generated from triplicates of data to calculate EC₅₀ values with Prism 5.0.

MTT test with the L-929 mouse fibroblasts

Growth inhibitory activity on L-929 mouse fibroblasts were determined after incubation with serial dilutions of the compounds for 5 d using an MTT assay. ^{27, 28} The IC₅₀ was estimated from the concentration dependent activity curves.

Scheme 1.

Synthetic strategy and key fragments for the synthesis of chondramide A analogues 2b–k.