Sesquiterpenes from the Brazilian Red Alga Laurencia dendroidea J. Agardh

Two new chamigrane sesquiterpenes 1–2 and three known compounds 3–5 were isolated from a lipophilic extract of the red alga Laurencia dendroidea collected from the Southeastern Brazilian coast. Dendroidone (1) and dendroidiol (2) were isolated from samples collected at Biscaia Inlet, Angra dos Reis, Rio de Janeiro and at Manguinhos Beach, Serra, Espírito Santo, respectively. Debromoelatol (3), obtusane (4) and (1S*,2S*,3S*,5S*,8S*,9S*)-2,3,5,9-tetramethyltricyclo[6.3.0.01.5]undecan-2-ol (5) were obtained from specimens collected at Vermelha Beach, Parati, Rio de Janeiro. The structures of new compounds were elucidated by extensive NMR (1H-, 13C-, COSY, HSQC, HMBC and NOESY) and high resolution mass spectrometry analysis. Additionally, the absolute configuration of compound 2 was assigned by X-ray analysis. Full spectroscopic data is described for the first time for compound 3. Anti-inflammatory and antimycobacterial activities of compounds 2–5 were evaluated. Compounds 3–5 inhibited the release of inflammatory mediator NO while TNF-α levels were only affected by 3. All compounds tested displayed moderate antimycobacterial action.


Introduction
It is estimated that more than 700 compounds with unique structural features have already been isolated from red algae of genus Laurencia, family Rhodomelaceae, order Ceramiales [1,2], which occurs on temperate to tropical shores of the world inhabiting intertidal and subtidal areas [3]. This remarkable chemical diversity comprises sesquiterpenes, diterpenes, triterpenes and acetogenins, mainly halogenated [4][5][6][7][8][9]. Several studies suggest that in the marine environment these compounds have a role as chemical defenses against herbivores, fouling organisms and pathogens [10,11]. Despite the high number of isolated compounds, recent reports confirm the potential of Laurencia to produce unknown structures [12]. Chamigrane-type compounds are the main class of sesquiterpenes isolated [2], for which some interesting pharmacological actions are described, including antibacterial [13], cytotoxic [14], and antileishmanial [15]. In addition, L. undulata and L. snackeyi extracts exhibited anti-inflammatory action [16,17], and the tricyclic brominated diterpene neorogioltriol, isolated from L. glandulifera, displayed both in vivo and in vitro anti-inflammatory activity [18].
The pathogenesis of several diseases such as tuberculosis, caused mainly by Mycobacterium tuberculosis, is highly influenced by the inflammatory response. Some anti-inflammatory drugs are employed as an adjunctive therapy for tuberculosis [19]. Therefore, combined anti-inflammatory and antimycobacterial properties in a single compound or class of compounds could be relevant for the treatment of tuberculosis. Furthermore, the long duration of current therapy as well as the associated side effects often compromises its effectiveness and it is intimately linked to the emergence of drug resistance [20]. Thus, there is an urgent need for short and simple regimens, which are both effective and safe.
In our ongoing study on structurally diverse and biologically active compounds from Laurencia species, specimens of the Brazilian red alga L. dendroidea J. Agardh were collected from three different places along the southeastern coast, and extracted with dichloromethane. Herein we report the extraction, isolation and structure elucidation of compounds 1-5 ( Figure 1) along with anti-inflammatory and antimycobacterial activities of compounds 2-5.
Ring B was proposed based on correlations between methylenes protons at δ H 2.81 (H-2a) and δ H 1.76 (H-1b) on the COSY spectrum along with HMBC correlations from δ H 2.46 (H-5b) to δ C 197.6 (C-4) and from δ H 1.76 (H-1b) to δ C 48.7 (C-6). Based on the 13 C-NMR data, a chlorine atom was assigned to C-15 [22]. The Z configuration of C-3 double bond was proposed from the observation of correlation between H 2 -2/H-15 on NOESY spectrum. The relative configuration was determined by NOESY correlations and 1 H-1 H coupling constants. NOE correlations of H-14b to H-5a demonstrated that exomethylene and methylene CH 2 -5 groups were positioned on the same face. Moreover, halomethine proton H-10 displayed correlations to H-1a and H 3 -13 indicating that H-10 occupied an axial position. Based on the axial orientation of H-10 and the small coupling constant of the carbinolic H-9 (3 Hz) it was suggested that it was equatorial, therefore bromine and hydroxyl were in a cis configuration. Hence, the combined data established the structure of compound 1 as (Z)-10-bromo-15chloro-11,11-dimethyl-7-methylidenespiro [5.5]undec-3(15)-ene-4-one which represents a new chemical entity, which was trivially named dendroidone.  The relative configuration was determined on the basis of measured coupling constants and NOESY spectrum ( Figure 2). Taken together, the data suggested that the compound 2 represented a new chamigrane sesquiterpene with chlorohydrin function 4,10-dibromo-4-chloro-3,11,11-trimethyl-7methylidenespiro [5.5]undec-3,9-diol, for which the trivial name dendroidiol was proposed. Further X-ray crystallographic data of 2 confirmed the suggested structure and defined the absolute configuration as 3R, 4S, 6S, 9R, 10S as depicted ( Figure 2). The rings A (C1-C2-C3-C4-C5-C6) and B (C7-C8-C9-C10-C11-C12) adopted chair configuration, with ring-puckering parameters q 2 = 0.060(7)Å;  2 = 352(7)° and q 2 = 0.042(5) Å;  2 = 233(7)°, respectively [24]. Compound 3 was isolated as colorless oil. The 13 C-NMR spectra along with HSQC experiment revealed the presence of fifteen carbons distributed as five quaternary carbons, one methine, six methylenes and three methyls, including four olefinic carbons ( Table 1). The molecular formula C 15 H 23 OCl was deduced by NMR and EI-MS data. Like compounds 1 and 2, the 1 H-NMR spectrum displayed singlets of an exocyclic methylene group (δ H 5.01, 4.60). Additionally, it also displayed one carbinolic proton at δ H 3.76 and three quaternary methyls (δ H 0.84, 0.94, 1.70). Comparison of NMR spectra of compound 3 to 1 and 2 revealed that ring A differed only on bromine substitution at C-10. The correlations in the HMBC spectrum from methyl H 3 -15 to C-2, C-3 and C-4 indicated the second ring was similar to the described for elatol [15]. Thus, the present data suggested that compound 3 was debromoelatol previously isolated from L. obtusa [25]. Full spectroscopic data for compound 3 is described for the first time.
Compounds 2-5 were submitted to tests evaluating immunomodulatory and antimycobacterial actions ( Table 2). These two pharmacological approaches were supported by previous studies reporting immunomodulatory [26] and antimycobacterial activities [27] of Laurencia species and also due to the expertise of the group in these two areas. The in vitro anti-inflammatory potential of isolated compounds was evaluated in a preliminary study of immunomodulatory properties, which was assessed by their inhibitory effects on NO and TNF-α productions from LPS-activated RAW 264.7 macrophages. Compounds 3-5 inhibited NO release by stimulated macrophages, with IC 50 values ranging from 44.9 ± 3.0 to 74.6 ± 5.8 μM (Table 2). Compound 4 was significantly more active (p < 0.05) while compounds 3 and 5 displayed similar activity to the positive control L-NMMA (L-N-monomethyl-arginine), a selective iNOS synthase inhibitor (p > 0.05). TNF-α production was moderately inhibited by compound 3 (IC 50 133.8 ± 7.4 µM), however, the remaining compounds did not show promising effects.
In the second part of the preliminary pharmacological study, antimycobacterial activity of isolated sesquiterpenes was evaluated using rapidly-growing strain Mycobacterium bovis BCG (Table 2). In this test, the sesquiterpene 4 (IC 50 44.7 ± 4.0 μM) was the most active compound, but it was less effective than the positive control Rifampicin.
In order to determine whether there was any selectivity, cell viability was assessed by lactate dehydrogenase (LDH) release (Table 2). Compound 4 was considered only moderately toxic while compounds 2, 3 and 5 showed no toxicity whatsoever.

General Procedures
Optical rotations were measured on a Perkin Elmer model 341LC polarimeter using a Na lamp at 20 °C. IR spectra were obtained with a Perkin Elmer spectrum one FT-IR. 1 H-NMR, 13 C-NMR, DEPT-135, COSY, HSQC, HMBC and NOESY spectra were measured employing a Bruker Avance III instrument operating at 500 MHz for 1 H-NMR and at 125 MHz for 13 C NMR in CDCl 3 . EI-MS spectra were obtained with a Shimadzu GCMSQP-2010 Plus. HR-APCI-MS spectra were recorded on a MicrOTOF (Bruker Daltonics, Billerica, MA, USA) mass spectrometer. HR-ESI-MS spectra were recorded on an UltrOTOF (Bruker Daltonics) mass spectrometer. Column chromatography was performed with Silicycle SiliaFlash F60 (230-400 mesh) silica and Sephadex LH-20 (Fluka, Steinheim, Germany). Thin layer chromatography was carried out with silica gel GF 254 plates. The spray reagent was a solution of 2% of Ce(SO 4 ) 2 in H 2 SO 4 . HPLC separations were performed with a Shimadzu instrument equipped with an LC-6AD pump, CBM-20A and SPD-20AV detector using a Shim-pack Prep-ODS, 250 × 20 mm, 5 μm column.

Plant Material
The red seaweed Laurencia dendroidea J. Agardh (Rhodomelaceae, Ceramiales) was collected from three distinct areas of the Southeastern Brazilian coast: Biscaia inlet -Angra dos Reis -Rio de Janeiro state (

X-ray Crystallography
X-ray diffraction data was carried out in Nonius Kappa CCD diffractometer at room temperature with radiation MoKα. The collect was performed utilizing Collect software [28] and the data was reduced with EvallCCD [24]. The structure was solved by direct methods and refined by full-matrix least squares on F2 with SHELX-97 package [29]. The positions of hydrogen atoms were generated geometrically and refined according to a riding model. All non-hydrogen atoms were refined anisotropically. The supplementary crystallographic data for 2 reported in this paper have been deposited at the Cambridge Crystallographic Data Center, under the reference number CCDC 950138. Copies of the data can be obtained, free of charge, on application to the Director, CCDC, 12 Union Road, Cambridge CB2 1EZ, UK, fax: +44 1223 336033 or data_request@ccdc.cam.ac.uk.

Antimycobacterial Activity
Samples were evaluated using a tetrazole salt assay to measure mycobacterial growth in liquid medium [31]. Initially, a suspension of Mycobacterium bovis BCG strain Moreau was grown in Middlebrook 7H9 medium supplemented with 0.05% Tween 80 and ADC. At a middle logarithmic growth phase, the bacterial suspension was diluted to obtain a concentration of 2 × 10 7 CFU/mL, and 50 µL of the resulting suspension was plated in a 96-well plate (1 × 10 6 CFU/well) and supplemented with 50 µL of each sample in three concentrations. The sealed plate was incubated at 37 °C and 5% CO 2 for 7 days. After this period, 10 µL of tetrazolium salt (MTT: 3-[4,5-dimethylthiazol-2-yl]-2,5diphenyltetrazole, Sigma-Aldrich, St. Louis, MO, USA -5 mg/mL in sterile PBS) was added. After 3 h of incubation, the cells were lysed through the treatment with 100 µL of lyses buffer (20% w/v SDS/ 50% DMF -dimethylformamide in distilled water -pH 4.7). The plate was incubated overnight and measured using a spectrophotometer at 570 nm. As a positive control, a bacterial suspension treated with the standard antimycobacterial drug rifampicin (Sigma-Aldrich-95% purity) at concentrations of 0.0011, 0.0033, 0.01 and 0.03 µg/mL, was used. As a negative control, an untreated bacterial suspension was employed. The test was performed in triplicate and the mean value and standard deviation were calculated.

Determination of Nitric Oxide and TNF-α Production by the RAW 264.7 Macrophage
The murine peritoneal macrophage cell line RAW 264.7 was obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA) and grown at 37 °C and 5% CO 2 in DMEM F-12 that was supplemented with 10% FCS and gentamicin (50 µg/mL). RAW 264.7 cells (1 × 10 5 cells/well) were seeded in flat bottom 96-well tissue culture plates (Corning Inc., Corning, NY, USA) in the presence or absence of various concentrations of the samples (100, 20 and 4 µg/mL) and/or LPS (Escherichia coli 055:B5; Sigma-Aldrich). After a 24 h incubation period, culture supernatants were collected for NO and TNF-α assays. Nitrite, a stable NO metabolite, was determined by using the Griess test [32]. As a positive control of inhibitory activity, intact, untreated macrophages were used. As a negative control, macrophages stimulated with 1 µg/mL LPS were used. A nitric oxide inhibitor, L-NMMA (Sigma-Aldrich -98% purity), was also used as a positive control at 20 µg/mL, inhibiting 59.22% ± 2.96% of the NO production. TNF-α was measured by an L929 fibroblast bioassay. This assay system uses murine L929 cells sensitive to TNF-α. For this, murine fibroblast cell line L929 cells (ATCC) (2 × 10 5 cells/well) were seeded in flat bottom 96-well tissue culture plates (Corning Inc.) 24 h before of being inoculated with macrophage culture supernatant and actinomycin D (2 µg/mL) added. After 24 h of incubation with macrophage culture supernatant, L929 viability was assayed by MTT [3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide] method [33]. The cytokine levels were calculated by using a purified recombinant mouse cytokine to obtain a standard curve that correlates cellular viability and TNF-α concentration.

Lactate Dehydrogenase Cytotoxicity Assay
LDH assay was used to evaluate the toxicity of the studied samples towards macrophage cultures. The release of LDH (cytoplasmic enzyme lactate dehydrogenase) from RAW 264.7 cells treated with samples was determined using 50 µL of cell culture supernatant collected 24 h after the treatment, as described in the previous section [34]. The LDH release, which represents an indirect indication of cytotoxicity, was determined colorimetrically using a commercial kit (Doles Reagentes e Equipamentos para Laboratorios Ltda., Goiânia, Brazil). The specific release was calculated as a percentage of the controls: non-treated macrophages as the negative control (O.D. 0.249, cytotoxicity 1.99% ± 0.62%) and 1% Triton X-100 (Vetec Chem., Duque de Caxias, Brazil) detergent treated macrophages as the positive control (O.D. 1.278, cytotoxicity 99.95% ± 2.26%). Final concentrations of DMSO, used as the carrier solvent for the samples, were tested in parallel as a control. Cytotoxicity was shown as percentage of controls. Tests were performed in triplicate and the mean value and standard deviation were calculated.

Conclusions
In conclusion, the present study resulted in the isolation of five sesquiterpenes from Brazilian specimens of L. dendroidea. Compound 1 represents a new compound with a chloroenone group while compound 2 displayed a chlorohydrin function. Full spectroscopic data is described for the first time for compound 3. Moreover, compound 4 significantly suppressed NO production in LPS-stimulated RAW 264.7 macrophages and also displayed antimycobacterial action against M. bovis BCG. Thus, present data showed that L. dendroidea is a promising source of immunomodulatory and antimycobacterial drugs.