Structure of debromo-carteramine A, a novel bromopyrrole alkaloid from the Mediterranean sponge Axinella verrucosa

The butanol extract of the Mediterranean sponge Axinella verrucosa was fractionated via a bioactivity-guided procedure based on an antibacterial assay. This approach led to the isolation of the novel alkaloid debromo-carteramine A 2 co-occurring with known hymenidin 1 , the main antibacterial agent of the extract, and related inactive carteramine A 3 .


Introduction
][3][4][5][6][7] Historically, these compounds, which have been found only in the marine environment to date, have attracted the attention of natural product chemists because of their structural complexity.Today, the interest is sparked not only by the wide structural diversity, ranging from simple molecules like oroidin 8,9 to complex structures like palau'amine 10,11 and stylissadine A and B, 12 but also by their organic synthesis, 13 bio-synthesis, 14 and pharmacological activities. 6,7,159][20] The crude extracts of different Axinella species have also been reported to inhibit the growth of several environmental bacteria. 21sing a bioassay-guided isolation approach, we have analyzed the butanol extract of a sample of A. verrucosa collected during the spring 2008 off Massa Lubrense, Bay of Naples.The extract showed strong antibacterial activity against the marine bacterium Lysinibacillus sp.It was found that the main component of the extract, hymenidin 1 (Figure 1), previously isolated from Hymeniacidon sponge, 22 was also the main compound responsible of the observed antimicrobial activity.Analysis of the fractions exhibiting a weak/moderate activity led to the isolation of a novel alkaloid of the palau'amine class, debromo-carteramine A 2 (Figure 1).Interestingly, the same fraction also contained the known related compound carteramine A 3 (Figure 1), recently isolated from Stylissa carteri 23 and identical with the compound named tetrabromostyloguanidine reported at the same time from Stylissa caribica. 24Carteramine A did not inhibit the bacterium of choice.In this paper, we report the isolation and structure elucidation of the novel compound 2.

Results and Discussion
An aliquot (185 mg) of the butanolic soluble portion (690 mg) of the acetone extract of the sponge (5.6 g, dry weight after extraction) was submitted to molecular exclusion chromatography (Sephadex LH-20, CHCl3/MeOH, 1:1) to give two distinct sets of active fractions.The fraction eluted first contained pure hymenidin 1 whereas the second fraction was a mixture that was further purified by reverse-phase HPLC with a linear gradient system to give a pure active compound, debromo-carteramine A (2, 4.0 mg), and the inactive related carteramine A (3, 3.7 mg).][24] Debromo-carteramine A 2 showed a sodiated-molecular peak at m/z 746.9163 in the HRESIMS spectrum, which indicated the molecular formula C22H22ClBr3N10O3 differing from that of the co-occurring 3 in the presence of a hydrogen atom instead of a bromine.Compound 2 displayed in the 13   In the 1 H NMR spectrum of 2, an additional 1H singlet at δ 7.12 (H-2) was present with respect to 3 whereas the remaining part of the spectrum was almost identical.This signal was attributed to an isolated proton in the α-position of a pyrrole ring thus indicating, in agreement with the molecular formula, that 2 lacked the bromine substituent at C-2 with respect to 3. Further evidence was obtained from the 13 C NMR spectrum of 2 containing the CH sp 2 signal in the place of a C sp 2 signal (C-2).The presence of the chlorinated 7-azabicyclo [3.3.0.octane ring system characterizing all members of the palau'amine class was also evident by analysis of the 1 H-1 H COSY spectrum.In fact the typical spin-system from H2-13 to H2-19 was easily detected.Analysis of 2D-NMR experiments ( 1 H-1 H COSY, HSQC, and HMBC) allowed complete proton and carbon assignments as reported in Table 1 and confirmed that 2 was the 2-debromoderivative of carteramine A 23 (≡ tetrabromostyloguanidine). 24 ISSN 1551-7012 Page 236  ARKAT USA, Inc.
The relative stereochemistry of the eight chiral centres of 2 was suggested to be the same as 3 by comparison of their 13 C-NMR data 23 and was further supported by analysis of a series of NOE difference and NOESY experiments (Figure 2) recorded in CD3OD and d6-DMSO, respectively.Analogous to the data reported for carteramine A, 23  Thus 2-debromo-carteramine A 2 exhibited the same relative configuration as carteramine A 23 (≡ tetrabromostyloguanidine), 24 the stereochemistry of which was determined by NOESY analysis 26 and computational methods, 24 and differed from those reported for palau'amine and related compounds. 10,11However, in both these recent papers 23,24 the relative configuration of palau'amine congeners was suggested to need revision.
Compounds 1 and 2 showed activity against the environmental marine bacterium Lysinibacillus sp.ESY 9 (GenBank accession no.GU059941) in the disc diffusion assay with an inhibition zone diameter of 12.5 mm at a concentration of 100 µg/disc and of 8 mm at a concentration of 200 µg/disc, respectively.It is interesting to note that co-occurring carteramine A (3) was found to be inactive in the same test suggesting that free C-2 position is an essential requirement for this activity.

Experimental Section
General.Optical rotations were measured on a JASCO DIP 370 digital polarimeter.The UV spectra were recorded on JASCO 710 spectropolarimeter. 1 H and 13 C NMR spectra were recorded on DRX 600, AVANCE 400, and DPX 300 MHz Bruker spectrometers in CD3OD and in d6-DMSO, with chemical shifts reported in ppm referred to CH3OH (δ 3.34 for proton and δ 49.9 for carbon) and to DMSO (δ 2.54 for proton and δ 40.5 for carbon), respectively, as internal standards.ESIMS and HRESIMS were measured on a Micromass Q-TOF Micro TM coupled with a HPLC Waters Alliance 2695.The instrument was calibrated by using a PEG mixture from 200 to 1000 MW (resolution specification 5000 FWHM, deviation <5 ppm RMS in presence of a known lock mass).Silica gel and exclusion chromatography were performed using precoated Merck F254 plates and Sephadex TM LH-20 (Amersham Biosciences), respectively.HPLC purifications were carried out on a Thermo Electron chromatograph coupled with P4000 pumps and a UV2000 double wavelength detector.

Collection and extraction of the animal material
Specimens of A. verrucosa were collected by Scuba during the Spring 2008 off Massa Lubrense, Bay of Naples at 25 m depth.The biological material was immediately transferred to the ICB laboratory, where it was cleaned from epibionts, rinsed once and frozen at -20 ºC, until its extraction.The frozen material was cut into pieces of about 1 cm 3 and extracted with acetone (250 mL × 3) under grinding and sonication.The three phases were combined after filtration and the organic solvent was removed under reduced pressure.The residual water was partitioned three times with Et2O and subsequently with n-butanol.The combined Et2O phases (461 mg) and the n-butanol (690 mg) phase were dried under reduced pressure to give the initial two extracts.

Purification of compounds
An aliquot of n-butanolic extract (185 mg) was fractionated on a Sephadex LH-20 column.The collected fractions were re-combined based on their TLC pattern resulting in 17 fractions.All fractions were tested for their antibacterial activity, observing two active fractions, A (72 mg) and B (9 mg), which were further analyzed by 1 H NMR. Fraction A contained compound 1 pure, whereas fraction B resulted a mixture, which was purified on reverse-phase HPLC (Phenomenex: Kromasil 5µ C18, 250x10 mm, 40 min gradient from 50% to 100% CH3OH in H2O with 0.1% of TFA, flow 2 mL/min, UV detector) to obtain compounds 2 and 3. Compound 1.Light yellow powder; 1

Antibacterial assays
Disc diffusion assays were performed using the environmental marine bacterium Lysinibacillus sp.strain ESY 9 (GenBank accession no.GU059941).The bacterium was isolated on LB plates (10 g L -1 Tryptone, 5 g L -1 Yeast Extract, 5 g L -1 NaCl, 15 g L -1 Bactoagar) from sediment collected at 30 m depth close to the sponge A. verrucosa at the reef in front of Sdot Yam, Israel.Test extracts were transferred to blank paper discs (diameter 6 mm) and left until the solvent completely evaporated.For the assay 250 µL of an overnight culture grown in liquid LB medium at 30 ºC were plated onto a LB plates until the surface was dry.The paper discs with the applied extracts were then transferred to the seeded plates.Plates were wrapped with Parafilm and incubated for one day at 30 ºC after which the diameter of inhibition zones, visible as clear zones around the paper discs, was measured to the next half millimeter.Solvent controls were always run in parallel and never showed inhibition zones.Assays were performed for the initial Et2O (dissolved in chloroform) and n-butanol (dissolved in MeOH) extracts, all Sephadex LH-20 fractions of the n-butanol, and all reverse-phase HPLC fractions of the second active Sephadex LH-20 fraction including the baseline collection and the column washes.