Cyclopsammocinamides A and B, Enantiomeric Cyclic Peptides of Cyclocinamide A, from the Marine Sponge Psammocinia sp.

Abstract

LC-MS and molecular networking analyses of the extract of the marine sponge Psammocinia sp. indicated the presence of two new compounds with multiple halogens. LC-MS-guided isolation yielded cyclic peptides, cyclopsammocinamides A (1) and B (2), in an enantiomeric relationship to cyclocinamide A (3). Planar structures of 1 and 2 were elucidated by NMR and mass spectroscopic analyses and the absolute configurations of the amino acid residues were determined using Marfey’s method with their acid hydrolysates. The sponge extract exhibited cytotoxicity and the bioassay-guided isolation afforded a dimeric dilactone macrolide, swinholide A, as the cytotoxic compound.

Introduction

In 1997, cyclocinamide A (3) was isolated from the marine sponge Psammocinia sp. obtained from Papua New Guinea by Crews and colleagues.¹⁾ It had a βαβα-14-membered cyclic tetrapeptide core structure with 5-bromotryptophane and 1-methyl-3-chloropyrrole moieties. The absolute configuration was shown to be 7S and 14S by chiral TLC using its hydrolysate, whereas the 4S and 11S configurations were presumed based on the biogenetic consideration. To determine the correct structure, the 4R,7S,11R,14S²⁾ and 4R,7S,11S,14S³⁾ isomers were synthesized, but their NMR data were not identical. In 2008, Crews and colleagues reisolated 3 and reported its absolute configuration as all-S using Marfey’s method.⁴⁾ In 2012, to examine the proposed structure, Konopelski and colleagues synthesized the all-S,⁵⁾ 4S,7S,11R,14S,⁵⁾ and 4S,7R,11S,14R⁶⁾ isomers, but their data were inconsistent with the natural product. In 2018, Konopelski and colleagues applied the DP4 probability method to the flexible cyclic peptide with four separated stereocenters and suggested that the correct configuration is 4S,7R,11R,14S. They synthesized the isomer, and the NMR data of the synthetic and natural compounds matched completely, and the specific rotation value showed the same sign.⁷⁾ Thus, the absolute configuration of 3 was established. In 2007, Ireland and colleagues isolated a congener, cyclocinamide B, from the marine sponge Corticicum sp. The planar structure was proposed to be the same as 3 but with additional chlorine at C-36, and the absolute configuration was assigned to be 4S,7R,11S,14R using Marfey’s method.⁸⁾ In 2015, the compound with the proposed configuration was synthesized, but the NMR data were not identical to those of the natural compound.⁶⁾

During our search for biologically active compounds from marine sponges, the extract of the sponge Psammocinia sp. collected in Indonesia exhibited potent cytotoxic activity against HCT-116 cells (6% viability at 50 µg/mL). LC-MS profiling of the active fraction, followed by MS/MS-based molecular networking analysis,⁹⁾ revealed the presence of two congeners with multiple halogens and unique mass spectra. These two compounds were isolated by LC-MS, and their structures were studied.

Results and Discussion

From LC-MS profiling of the biologically active extracts of marine sponges, the EtOH extract of the marine sponge Psammocinia sp. exhibited two peaks with multiple halogens. The 90% MeOH-soluble fraction obtained from the EtOAc-soluble fraction of the extract was analyzed by MS/MS-based molecular networking. The results revealed that these compounds, cyclopsammocinamides A (1) and B (2), were structurally related, which was confirmed by their fragmentation patterns (Figs. 1, 2). The molecular formulas of 1 and 2 were found to be C₂₈H₃₁BrClN₉O₈ and C₂₈H₃₀BrCl₂N₉O₈, respectively, and they were believed to be new compounds.

Fig. 1. Chemical Structures of Compounds 1–3

Fig. 2. MS/MS-Based Molecular Networking Analysis of the 90% MeOH Fraction of the EtOH Extract of the Marine Sponge Psammocinia sp. and MS/MS Spectra of 1 and 2

(A) Green nodes are for 1 and 2, and the yellow nodes are for preswinholide A and swinholide A. (B) MS/MS spectra of 1 and 2 in positive and negative modes showing similar fragmentation patterns.

The ¹H- and ¹³C-NMR spectra of 1 and 2 measured in dimethyl sulfoxide (DMSO)-d₆ (Table 1, Supplementary Figs. S1, S2, S6, S7) revealed their peptide nature, and analysis of two dimensional (2D) NMR spectra including correlation spectroscopy (COSY), heteronuclear single quantum coherence (HSQC), and heteronuclear multiple bond connectivity (HMBC) (Fig. 3) revealed six substructures, isoserine (iSer), 5-bromotryptophan (5-BrTrp), 2,3-diaminopropionic acid (DAP), asparagine (Asn), glycine (Gly), and a 2,4-disubstituted pyrrole. The planar structures of 1 and 2 were almost the same as that of 3, except for the chlorinated pyrrole moiety. The HMBC correlations from H₂-30 and NH-31 to C-32 of 1 indicated that N-31 of the glycine residue was connected to C-32, indicating two possibilities for the terminal structure such as the 2-substituted 4-chloropyrrole and 4-substituted 2-chloropyrrole moieties. Although the HMBC correlations from the protons [δ 6.84 (t, J = 1.9 Hz) and 6.97 (dd, J = 2.7 and 1.9 Hz)] of the pyrrole moiety to C-32 indicated the 4-substituted 2-chloropyrrole moiety as the preferable structure, the observation of a four-bond correlation H-36/C-32 was also reported for 3,¹⁾ containing a 2-substituted 1-methyl-4-chloropyrrole moiety. To determine the substituted position of the pyrrole moiety in 1, DFT calculation of the ¹³C-NMR data was conducted for models A and B (Fig. 4). The correlation between the calculated and experimental data for model A (R² = 0.9977) was higher than that for model B (R² = 0.9902) (Fig. 4). These results indicated that a 2-substituted 4-chloropyrrole moiety was connected to C-32 in 1. The molecular formula of 2 suggested the presence of another chlorine in 1, and the HMBC correlations showed that it was attached to C-36.

Table 1. ¹H- and ¹³C-NMR Spectroscopic Data (DMSO-d₆) of 1 and 2

	1			2
Pos.	δC, mult.	δH, mult. (J in Hz)	HMBC	δC, mult.	δH, mult. (J in Hz)	HMBC
1	170.6, C			170.6, C
2		7.14, t (5.6)	1, 3, 4		7.14, t (5.5)	1, 3
3	42.7, CH2	3.42, m	4	42.7, CH2	3.41, m	4, 5
		3.48, m	4, 5		3.48, m	4, 5
4	69.8, CH	4.04, br t (4.2)	3, 5	69.8, CH	4.05, br s	5
5	171.1, C			171.1, C
6		7.89, d (9.8)	5, 7, 18		7.89, d (9.2)	5, 7, 8
7	53.4, CH	4.57, m	5, 8, 18, 19	53.4, CH	4.57, m	5, 8, 18, 19
8	172.5, C			172.6, C
9		7.89, br s	7, 8, 10		7.90, br s	7, 8, 10
10	40.3, CH2	3.36, m	8	40.4, CH2	3.33, m	8
		3.40, m	8		3.39, m	8
11	54.3, CH	4.32, dt (7.9, 3.5)	10, 12, 29	54.4, CH	4.31, dt (7.8, 3.5)	10, 12, 29
12	168.8, C			168.7, C
13		7.99, d (8.8)	12, 14, 15		7.99, br d (8.8)	12, 14
14	49.6, CH	4.56, m	1, 12, 15, 16	49.6, CH	4.56, m	12, 15, 16
15	36.5, CH2	2.33, dd (15.5, 4.9)	1, 14, 16	36.4, CH2	2.33, dd (15.6, 4.9)	1, 14, 16
		2.45, dd (15.5, 7.1)	1, 14, 16		2.44, dd (15.6, 7.1)	1, 14, 16
16	172.0, C			172.0, C
17		6.82, br s	15, 16		6.82, br s
		7.27, br s	16		7.28, s	16
18	27.8, CH2	2.98, dd (14.5, 7.1)	7, 8, 19, 20, 27	27.8, CH2	2.98, dd (13.0, 5.7)	7, 8, 19, 27
		3.02, dd (14.5, 6.2)	7, 8, 19, 20, 27		3.02, dd (13.0, 4.6)	7, 8, 19, 27
19	109.6, C			109.5, C
20	125.3, CH	7.17, d (1.8)	18, 19, 22, 27	125.3, CH	7.16, d (2.0)	22, 27
21		11.07, d (1.8)	19, 20, 22, 27		11.08, d (2.0)	19, 20, 22, 27
22	134.8, C			134.8, C
23	113.4, CH	7.29, d (8.6)	25, 27	113.4, CH	7.29, d (8.4)	25, 27
24	123.4, CH	7.15, d (8.6)	22, 26	123.4, CH	7.16, dd (8.4, 2.0)	25, 26
25	111.1, C			111.1, C
26	120.7, CH	7.66, br s	19, 22, 24, 25	120.7, CH	7.65, d (2.0)	19, 22, 24, 25, 27
27	129.2, C			129.1, C
28		8.18, d (7.9)	10, 11, 29		8.10, d (7.8)	10, 29
29	169.2, C			169.0, C
30	42.3, CH2	3.83, dd (16.7, 6.0)	29, 32	42.2, CH2	3.84, dd (16.8, 6.1)	29, 32
		3.88, dd (16.7, 6.0)	29, 32		3.88, dd (16.8, 6.1)	29, 32
31		8.43, t (5.7)	30, 32		8.49, t (5.7)
32	160.2, C			159.5, C
33	125.6, C			124.5, C
34	109.7, CH	6.84, t (1.9)	32, 33, 36	110.3, CH	6.95, s	32, 33, 36
35	110.7, C			108.1, C
36	119.1, CH	6.97, dd (2.7, 1.9)	32, 33, 34	115.1, C
37		11.73, br t (2.7)			12.69, br s	33, 34, 36
4-OH		5.98, d (4.8)			5.99, d (4.8)	3, 4, 5

Fig. 3. COSY (Bold Lines) and Key HMBC (Blue Arrows) of 1 and 2

Fig. 4. Correlations of Experimental ¹³C Chemical Shifts (C-29–C-36) of 1 versus Calculated Shifts of Models A and B

Green and red values on the structures of models A and B are calculated ¹³C chemical shifts.

To determine the absolute configurations of the amino acid residues in 1 and 2, Marfey’s method¹⁰⁾ was applied to their acid hydrolysates. Comparing LC-MS chromatograms of the derivatives prepared from 1 and 2 with those of standard amino acids revealed the presence of D-Asp, D-iSer, L-DAP, and L-5-BrTrp in 1 and 2, indicating the 4R,7S,11S,14R configuration for both 1 and 2 (Supplementary Fig. S11). Interestingly, this result also revealed that 1 and 2 have an enantiomeric relationship with 3. Signs of the specific rotations of 1 and 2, [α]²⁵_D −15 (c 0.1, MeOH) and −3.3 (c 1.7, MeOH), respectively, are opposite to that of 3, [α]²⁷_D +29 (c 0.1, MeOH)¹⁾ or [α]²⁷_D +32 (MeOH),⁸⁾ which indicate the enantiomeric relationship.

Although the sponge extract exhibited potent cytotoxicity against HCT-116 colon cancer cells, neither 1 nor 2 exhibited any cytotoxicity, and the bioassay-guided isolation showed that swinholide A^11,12) was the cytotoxic compound in this sponge.

Experimental

General

Optical rotations were measured using a JASCO DIP-1000 polarimeter in MeOH, UV spectra using a JASCO V-550 spectrophotometer in 50% MeCN–H₂O, IR spectra using a PerkinElmer, Inc. Frontier Fourier transform (FT)-IR spectrophotometer, and ¹H- and ¹³C-NMR spectra using a Bruker Avance III 600 NMR spectrometer or Bruker Avance III HD 500 NMR spectrometer. Chemical shifts were referenced to the residual solvent peaks (δ_H 2.49 and δ_C 39.51 for DMSO-d₆). Liquid chromatography-tandem mass spectrometry (LC-MS/MS) experiments were performed using a Shimadzu Prominence HPLC system equipped with LC-20AD pumps and a DAD SPD-M20A detector connected to a Bruker amaZon speed mass spectrometer. High-resolution mass spectra were measured using a Waters Xevo G2-XS Qtof mass spectrometer or Bruker Daltonics K.K. (Tokyo, Japan). Electrospray ionization (ESI)-Q-q-TOF mass spectrometer Impact II KUP. The preparative MPLC was performed using a Biotage Isolera I. The HPLC system consisted of a Waters 515 HPLC pump and Waters 2489 UV/visible detector.

Animal Material

The sponge, Psammocinia sp., was collected by scuba at a depth of 10 m in North Sulawesi, Indonesia, in September 2014 and immediately soaked in EtOH. The sponge was identified by one of the authors (Y.I.) and a voucher specimen (14M203) was deposited at the Department of Natural Medicines, Graduate School of Pharmaceutical Sciences, Kumamoto University, Japan.

Extraction and Isolation

The sponge (1.34 kg, wet weight) was extracted with EtOH. After evaporation, the remaining residue was suspended in H₂O and extracted with EtOAc followed by n-BuOH. The EtOAc fraction (4.15 g) was partitioned between n-hexane and 90% MeOH–H₂O. The 90% MeOH–H₂O soluble fraction (3.51 g) was subjected to SiO₂ column chromatography eluted with a stepwise gradient using 5, 10, and 20% MeOH–CH₂Cl₂ and CH₂Cl₂–MeOH–H₂O (6 : 4 : 1) to yield 12 fractions (Frs. A1–A12). Monitoring by LC-MS revealed that Fr. A10 (38.7 mg) eluted with CH₂Cl₂–MeOH–H₂O (6 : 4 : 1) yielded compounds 1 (t_R, 4.1 min; m/z 736 [M + H]⁺) and 2 (t_R 4.3 min; m/z 770 [M + H]⁺). Fr. A10 was then subjected to HPLC (COSMOSIL 5C18-AR-II, 20 × 250 mm, Nacalai Tesque Inc.) with 35% MeCN–H₂O to give 1 (t_R 22 min, 2.31 mg) and 2 (t_R 31 min, 1.29 mg). Fr. A3 (980 mg) was subjected to medium pressure liquid chromatography (MPLC) (Purif ODS, 30 g, Size 60, Biotage Japan Ltd., Tokyo, Japan) with 10 (0–12 min), 20 (12–24 min), and 100% (24–36 min) MeCN–H₂O to yield swinholide A (307 mg).

Cyclopsammocinamide A (1)

Yellowish, amorphous solid; [α]²⁵_D−15 (c 0.1, MeOH); UV (50% MeCN–H₂O) λ_max (log ε) 274 (4.01) and 226 (4.45) nm; IR (film) ν_max 3280, 2966, 2925, 2856, 1654, 1525, 1434, 1330, 1231, 1109, 1023, 937, and 603 cm⁻¹; ¹H- and ¹³C-NMR, Table 1; HRESITOFMS m/z 734.1119 [M − H]⁻ (Calcd for C₂₈H₃₀⁷⁹Br³⁵ClN₉O₈, 734.1095).

Cyclopsammocinamide B (2)

Yellowish, amorphous solid; [α]²⁵_D −3.3 (c 1.7, MeOH); UV (50% MeCN–H₂O) λ_max (log ε) 277 (4.07), 227 (4.42), and 210 (4.34) nm; IR (film) ν_max 3282, 2954, 2925, 2855, 1659, 1531, 1436, 1333, 1233, 1109, 1024, 1000, and 600 cm⁻¹; ¹H- and ¹³C-NMR, Table 1; HRESITOFMS m/z 770.0851 [M + H]⁺ (Calcd for C₂₈H₃₁⁷⁹Br³⁵Cl₂N₉O₈, 770.0851).

LC-MS Conditions

HPLC: column, COSMOSIL 2.5C18-MS-II, 2.0 × 100 mm (Nacalai Tesque, Inc., Kyoto, Japan); solvent system, a linear gradient of two solvents, 0.1% AcOH–H₂O (solvent A) and 0.1% AcOH–CH₃CN (solvent B), 10–100% B for 10 min; flow rate, 0.3 mL/min. The mass spectra were detected in the positive and negative ionization modes.

Molecular Networking Analysis

For molecular networking, the MS/MS data were converted to mzML files using MSConvert. Then, they were uploaded to the publicly available Global Natural Product Social Molecular Networking platform (http://gnps.ucsd.edu, accession on May 27, 2022).⁹⁾ Cytoscape (3.7.1) was used for the visualization and analysis of the output molecular network.

Acid Hydrolysis and L-FDAA Derivatization of 1 and 2

Each compound (0.1 µmol of 1 or 2) was dissolved in 1 mL of 6 N HCl containing 0.1% phenol (w/v) in a sealed ampule, and the solution was heated at 110 °C for 4 h. After condensation, the dried residue was dissolved in 100 µL of H₂O and reacted with 100 µL of 1% solution of L-FDAA in acetone (w/v) and 300 µL of 1 M NaHCO₃ at 50 °C for 2 h.¹⁰⁾ The reaction mixture was condensed and dissolved in MeCN for analysis. Standard D- and L-amino acids were derivatized under the same conditions. The L-FDAA derivatives were analyzed by LC-MS using a COSMOSIL 2.5C18-MS-II column (2.0 × 100 mm) at 40 °C eluted with a gradient of two solvents, 0.1% AcOH–H₂O (solvent A) and 0.1% AcOH–CH₃CN (solvent B), at 0.3 mL/min. The gradient program was conducted as follows: 10% B for 10 min, 10–55% B for 30 min, and 55–100% B for 20 min. The retention times of the standard amino acids were as follows: L-Asp (17.6 min), D-Asp (19.9 min); L-iSer (19.1 min), D-iSer (18.7 min); L-DAP (8.0 min), D-DAP (10.5 min); L-5-BrTrp (31.6 min), D-5-BrTrp (33.5 min) (Supplementary Fig. S11).

Conformational Analysis and ¹³C Chemical Shift Calculations

Conformational searches and chemical shift calculations for models A and B were carried out using Spartan’18 software (ver. 1.4.4, Wavefunction Inc., Irvine, CA, U.S.A.) on a commercially available PC (operating system: Windows 10 Education 64-bit, CPU: 6-Core Core i7-8700k processor 3.70 GHz, RAM 64 GB). Stable conformers were initially searched up to 10 kcal/mol using the Merck molecular force field (MMFF).¹³⁾ The energy of suggested conformers was calculated via Hartree–Fock (HF)/3-21G, and conformers with 40 kcal/mol higher than the least energy conformer were eliminated. Shortlisted conformers were then optimized via Hartree–Fock (HF)/6-31G* and conformers with less than 1% Boltzmann population were deleted. Chemical shifts of stable conformers were calculated at the ωB97X-D/6-31G* level. The obtained chemical shifts for each conformer were weighted using a Boltzmann distribution to yield the calculated ¹³C chemical shifts.

Cytotoxicity Assay against HCT-116 Cells

The cytotoxicity test was performed as described before.¹⁴⁾ Briefly, cells were grown in Dulbecco’s modified Eagle’s medium (DMEM)/F12 (1 : 1) (GIBCO, U.S.A.) supplemented with 10% fetal bovine serum, penicillin (50 units/mL), and streptomycin (50 µg/mL) under a humidified atmosphere containing 5% CO₂ at 37 °C. Cells were seeded on 96-well microplates (3 × 10³ cells/well) and pre-cultured for one day. The medium was replaced with that containing test compounds at various concentrations, and the cells were further cultured at 37 °C for three days. Then, the medium was replaced with 50 µL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) solution (200 µg/mL in medium), and the cells were incubated under the same conditions for 3 h. After the addition of 200 µL of DMSO, the optical density at 570 nm was measured with a microplate reader.

Acknowledgments

This work was supported by a JSPS KAKENHI Grant Number JP26305005 (S.T.).

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

References

• 1) Clark W. D., Corbett T., Valeriote F., Crews P., J. Am. Chem. Soc., 119, 9285–9286 (1997).
• 2) Grieco P. A., Reilly M., Tetrahedron Lett., 39, 8925–8928 (1998).
• 3) Liu L., “Ph.D. Dissertation,” Wayne State University, Detroit, MI, U.S.A., 2002.
• 4) Rubio B. K., Robinson S. J., Avalos C. E., Valeriote F. A., de Voogd N. J., Crews P., J. Nat. Prod., 71, 1475–1478 (2008).
• 5) Garcia J. M., Curzon S. S., Watts K. R., Konopelski J. P., Org. Lett., 14, 2054–2057 (2012).
• 6) Curzon S. S., Garcia J. M., Konopelski J. P., Tetrahedron Lett., 56, 2991–2994 (2015).
• 7) Cooper J. K., Li K., Aubé J., Coppage D. A., Konopelski J. P., Org. Lett., 20, 4314–4317 (2018).
• 8) Laird D. W., Labarbera D. V., Feng X., Bugni T. S., Harper M. K., Ireland C. M., J. Nat. Prod., 70, 741–746 (2007).
• 9) Yang J. Y., Sanchez L. M., Rath C. M., Liu X., Boudreau P. D., Bruns N., Glukhov E., Wodtke A., de Felicio R., Fenner A., Wong W. R., Linington R. G., Zhang L., Debonsi H. M., Gerwick W. H., Dorrestein P. C., J. Nat. Prod., 76, 1686–1699 (2013).
• 10) Marfey P., Carlsberg Res. Commun., 49, 591–596 (1984).
• 11) Carmely S., Kashman Y., Tetrahedron Lett., 26, 511–514 (1985).
• 12) Kitagawa I., Kobayashi M., Katori T., Yamashita M., Tanaka J., Doi M., Ishida T., J. Am. Chem. Soc., 112, 3710–3712 (1990).
• 13) Halgren T. A., J. Comput. Chem., 17, 490–519 (1996).
• 14) Afifi A. H., El-Desoky A. H., Kato H., Mangindaan R. E. P., de Voogd N. J., Ammar N. M., Hifnawy M. S., Tsukamoto S., Tetrahedron Lett., 57, 1285–1288 (2016).