Rugulopteryx-Derived Spatane, Secospatane, Prenylcubebane and Prenylkelsoane Diterpenoids as Inhibitors of Nitric Oxide Production

This study aimed to evaluate the anti-inflammatory potential of the different classes of diterpenoids produced by algae of the genus Rugulopteryx. First, sixteen diterpenoids (1–16), including spatane, secospatane, prenylcubebane, and prenylkelsoane metabolites, were isolated from the extract of the alga Rugulopteryx okamurae collected at the southwestern Spanish coasts. Eight of the isolated diterpenoids are new compounds whose structures were determined by spectroscopic means: the spatanes okaspatols A-D (1–4); the secospatane rugukamural D (8); the prenylcubebanes okacubols A (13) and B (14); and okamurol A (16), which exhibits an unusual diterpenoid skeleton featuring a kelsoane-type tricyclic nucleus. Second, anti-inflammatory assays were performed on microglial cells Bv.2 and macrophage cells RAW 264.7. Compounds 1, 3, 6, 12, and 16 caused significant inhibition of the NO overproduction induced by LPS in Bv.2 cells, and compounds 3, 5, 12, 14, and 16 significantly decreased levels of NO in LPS-stimulated RAW 264.7 cells. The most active compound was okaspatol C (3), which completely suppressed the effects of LPS stimulation, both in Bv.2 and in RAW 264.7 cells.


Introduction
The genus Rugulopteryx of brown algae was described in 2006 by De Clerck et al. to accommodate the species Dictyota radicans, Dilophus suhrii, and Dilophus marginatus, which were removed from Dictyota and Dilophus genera and, consequently, renamed as Rugulopteryx radicans, R. shurii, and R. marginata [1]. Later, in 2009, the species known as Dilophus okamurae was also transferred to the genus Rugulopteryx [2]. Although these four species of macroalgae typically grow along several areas of the Pacific Ocean [3], in recent years, R. okamurae has become renowned for having severely invaded the coasts of the Strait of Gibraltar, which connect the Mediterranean Sea and the Atlantic Ocean [4].
It is noteworthy that secospatane diterpenoids have been exclusively described from algae of the genus Rugulopteryx. The only exception is a secospatane reported from an alga identified as Dictyota fenestrata [16], which on a chemotaxonomic basis, suggests that the classification of this species might need a revision. On the other hand, spatane diterpenoids have been isolated from other brown algae of the Family Dictyotaceae, such as D. fenestrata [16], Spatoglossum schmittii [17,18], S. howleii [18], and Stoechospermum marginatum [19][20][21][22][23]. More recently, four spatanes have also been described from soft corals [24][25][26]. Regarding the prenylcubebanes, besides Rugulopteryx algae, the gorgonians Anthogorgia sp. and Euplexaura sp. have yielded terpenoids displaying this skeleton, although the four compounds isolated from these gorgonians were named as serrulatane-type diterpenoids [27,28]. In addition, the species R. okamurae has been the source of a few unique compounds with unprecedented diterpenoid skeletons, such as the dictyterpenoids isolated from Japanese specimens [13] and rugukadiol A from Spanish specimens of the alga [15]. cleavage of the bond at C-4,C-10. A third characteristic group comprises four compounds exhibiting a prenylcubebane framework, characterized by possessing a fused 5/3/6 tricyclic ring system [5,7,10,14]. It is noteworthy that secospatane diterpenoids have been exclusively described from algae of the genus Rugulopteryx. The only exception is a secospatane reported from an alga identified as Dictyota fenestrata [16], which on a chemotaxonomic basis, suggests that the classification of this species might need a revision. On the other hand, spatane diterpenoids have been isolated from other brown algae of the Family Dictyotaceae, such as D. fenestrata [16], Spatoglossum schmittii [17,18], S. howleii [18], and Stoechospermum marginatum [19][20][21][22][23]. More recently, four spatanes have also been described from soft corals [24][25][26]. Regarding the prenylcubebanes, besides Rugulopteryx algae, the gorgonians Anthogorgia sp. and Euplexaura sp. have yielded terpenoids displaying this skeleton, although the four compounds isolated from these gorgonians were named as serrulatane-type diterpenoids [27,28]. In addition, the species R. okamurae has been the source of a few unique compounds with unprecedented diterpenoid skeletons, such as the dictyterpenoids isolated from Japanese specimens [13] and rugukadiol A from Spanish specimens of the alga [15]. The noticeable structural diversity and unusual frameworks of the diterpenoids produced by algae of the genus Rugulopteryx offer great opportunities in the search for leads for the development of novel molecules for biomedical applications. In this regard, we have recently reported the first account on the anti-inflammatory activity of compounds of the secospatane class, in particular, of ruguloptone A and also of the uncommon diterpenoid rugukadiol A obtained from R. okamurae collected in the Spanish coasts [15]. These results prompted us to investigate the properties of diterpenoids of the spatane and prenylcubebane classes, whose potential remained unexplored, and to test additional secospatane metabolites. Herein we describe the isolation from R. okamurae of further sixteen diterpenoids. The isolated metabolites include seven spatanes (1-7), five secospatanes (8)(9)(10)(11)(12), and three prenylcubebanes (13)(14)(15), among which okaspatols A-D (1-4), rugukamural D (8), okacubols A (13), and B (14) are new compounds. In addition, the new okamurol A (16), which possesses an unusual kelsoane-type tricyclic nucleus, was also isolated. The capability of diterpenoids of the four structural classes to modulate the production of inflammatory mediators, such as nitric oxide (NO), was assayed. In particular, the anti-inflammatory effects were tested in a classic proinflammatory environment by LPS-stimulated microglial cells Bv.2 and macrophage cells RAW 264.7.

Results and Discussion
Fresh specimens of R. okamurae were treated with acetone/MeOH, and after evaporation of the solvent, the aqueous residue was extracted with diethyl ether. Column chromatography (CC) of the extract, followed by repeated CC and HPLC purification of selected fractions, led to the isolation of compounds 1-16 ( Figure 2). The noticeable structural diversity and unusual frameworks of the diterpenoids produced by algae of the genus Rugulopteryx offer great opportunities in the search for leads for the development of novel molecules for biomedical applications. In this regard, we have recently reported the first account on the anti-inflammatory activity of compounds of the secospatane class, in particular, of ruguloptone A and also of the uncommon diterpenoid rugukadiol A obtained from R. okamurae collected in the Spanish coasts [15]. These results prompted us to investigate the properties of diterpenoids of the spatane and prenylcubebane classes, whose potential remained unexplored, and to test additional secospatane metabolites. Herein we describe the isolation from R. okamurae of further sixteen diterpenoids. The isolated metabolites include seven spatanes (1-7), five secospatanes (8)(9)(10)(11)(12), and three prenylcubebanes (13)(14)(15), among which okaspatols A-D (1-4), rugukamural D (8), okacubols A (13), and B (14) are new compounds. In addition, the new okamurol A (16), which possesses an unusual kelsoane-type tricyclic nucleus, was also isolated. The capability of diterpenoids of the four structural classes to modulate the production of inflammatory mediators, such as nitric oxide (NO), was assayed. In particular, the anti-inflammatory effects were tested in a classic proinflammatory environment by LPS-stimulated microglial cells Bv.2 and macrophage cells RAW 264.7.

Results and Discussion
Fresh specimens of R. okamurae were treated with acetone/MeOH, and after evaporation of the solvent, the aqueous residue was extracted with diethyl ether. Column chromatography (CC) of the extract, followed by repeated CC and HPLC purification of selected fractions, led to the isolation of compounds 1-16 ( Figure 2).

Spatane Diterpenoids 1-7
Okaspatol A (1) possesses the molecular formula C22H34O3, determined by HRESIMS. The IR spectrum showed absorption bands at 3387 and 1711 cm −1 due to hydroxy and carbonyl functions, respectively. In particular, the presence of an acetate group was defined from the 1 H NMR signal at δ 2.03 (3H, s) and the 13 C NMR signals at δ 172.6 (-COO-) and δ 21.2 (CH3-) (Table 1). Another five methyl groups in the 1 H NMR spectrum (δH 1.76, 1.67, 1.63, 0.94, 0.82) and the remaining twenty signals of the 13 C NMR spectrum suggested a diterpene framework for compound 1.
The relative configuration of the molecule was defined from the NOESY data ( Figure 3b). The correlations H-8/H-2b, Me-12, and H-9/H-6a supported the cis-anti-cis arrangement of the 5/4/5 tricyclic ring system, exhibiting H-8 and Me-12 on one face of the molecule and H-9 and the hydroxy group on the opposite one. Following this, the NOESY correlations of H-8 with H-7 and Me-11 defined the orientation of the side chain and of Me-11. The α orientation of the acetate group at C-5 was assigned on the basis of the multiplicity of H-5, a doublet with J = 4.7 Hz, similar to that reported for spatol, whose stereochemistry was secured by X-ray diffraction analysis [17], for its acetate [18] and related compounds [18,19]. Moreover, when the acetoxy group at C-5 was β-oriented, H-5 has been described as a doublet of doublets with J = 10 and 8 Hz [19]. The NOESY correlations of H-5 with both methylene protons at C-6, while H-7 was only correlated with H-6b, were also consistent with the opposite orientation between H-5 and H-7. The Z configuration of the double bond at C-13,C-15 was defined by the NOESY correlation H-15/Me-14 and the 13 C NMR chemical shift of Me-14 at δ C 24.0 [29]. multiplicity of H-5, a doublet with J = 4.7 Hz, similar to that reported for spatol, whose stereochemistry was secured by X-ray diffraction analysis [17], for its acetate [18] and related compounds [18,19]. Moreover, when the acetoxy group at C-5 was β-oriented, H-5 has been described as a doublet of doublets with J = 10 and 8 Hz [19]. The NOESY correlations of H-5 with both methylene protons at C-6, while H-7 was only correlated with H-6b, were also consistent with the opposite orientation between H-5 and H-7. The Z configuration of the double bond at C-13,C-15 was defined by the NOESY correlation H-15/Me-14 and the 13 C NMR chemical shift of Me-14 at δC 24.0 [29].
(a) (b) Okaspatol B (2) possesses the molecular formula C20H32O2, determined by HRESIMS. The NMR spectra (Table 1) were related to those of compound 1, although diagnostic differences were observed, including the absence of the signals due to the acetate group, the shielding of the oxymethine (δC 75.7/δH 4.17), and the presence of an olefinic methylene (δC 108.9/δH 4.82 and 4.76). The analysis of COSY, HSQC, and HMBC spectra confirmed the spatane framework of compound 2 ( Figure S29). The presence of a tertiary hydroxy group at C-10 was supported by the HMBC correlations of the oxygenated carbon at δC 82.4 (C-10) with H-8, H-9, and Me-12, while the location of a secondary hydroxy group at C-5 was supported by the HMBC correlations of the oxymethine proton (δH 4.17) with C-7, C-8, and C-10. Differing from 1, the side chain of 2 contained a double bond at C-13,C-14, as indicated by the HMBC correlation of H-7 with the olefinic methylene carbon. The NOESY correlations ( Figure S30) and the multiplicity of H-5 (d, J = 4.5 Hz) indicated that compound 2 possessed the same relative configuration as 1 in all its chiral centers.
The molecular formula C20H32O2 was determined by HRESIMS analysis of okaspatol C (3). Upon comparison with compound 1, in the 1 H NMR spectrum of 3 ( Table 2) was evident the absence of the signals of the acetate, the oxymethine, and the Me-12 groups, showing, in turn, the signals of an AB system at δH 3.63 (d, J = 10.8 Hz, H-12a) and 3.43 (d, J = 10.8 Hz, H-12b) attributable to an isolated oxymethylene. These data suggested that compound 3 was a spatane diterpenoid bearing a primary hydroxy group at C-12. This proposal was confirmed by the HMBC correlations of the oxymethylene protons with C-5, C-8, and C-10 of the spatane tricyclic nucleus ( Figure S29). The NOESY correlations ( Figure S30) indicated that the relative configuration of compound 3 was identical to that of 1 and 2.
Okaspatol D (4) possessed the molecular formula C20H32O, determined by HRESIMS. The NMR spectra (Table 2) were related to those of the spatanes discussed above. However, the 1 H NMR spectrum of 4 did not show any signal attributable to protons geminal to an oxygenated function. In fact, the 13 C NMR spectrum of 4 only showed a carbon linked to oxygenated function at δC 82.2, which was identified as C-10 from its HMBC correla- Okaspatol B (2) possesses the molecular formula C 20 H 32 O 2, determined by HRESIMS. The NMR spectra (Table 1) were related to those of compound 1, although diagnostic differences were observed, including the absence of the signals due to the acetate group, the shielding of the oxymethine (δ C 75.7/δ H 4.17), and the presence of an olefinic methylene (δ C 108.9/δ H 4.82 and 4.76). The analysis of COSY, HSQC, and HMBC spectra confirmed the spatane framework of compound 2 ( Figure S29). The presence of a tertiary hydroxy group at C-10 was supported by the HMBC correlations of the oxygenated carbon at δ C 82.4 (C-10) with H-8, H-9, and Me-12, while the location of a secondary hydroxy group at C-5 was supported by the HMBC correlations of the oxymethine proton (δ H 4.17) with C-7, C-8, and C-10. Differing from 1, the side chain of 2 contained a double bond at C-13,C-14, as indicated by the HMBC correlation of H-7 with the olefinic methylene carbon. The NOESY correlations ( Figure S30) and the multiplicity of H-5 (d, J = 4.5 Hz) indicated that compound 2 possessed the same relative configuration as 1 in all its chiral centers.
The molecular formula C 20 H 32 O 2 was determined by HRESIMS analysis of okaspatol C (3). Upon comparison with compound 1, in the 1 H NMR spectrum of 3 (Table 2) was evident the absence of the signals of the acetate, the oxymethine, and the Me-12 groups, showing, in turn, the signals of an AB system at δ H 3.63 (d, J = 10.8 Hz, H-12a) and 3.43 (d, J = 10.8 Hz, H-12b) attributable to an isolated oxymethylene. These data suggested that compound 3 was a spatane diterpenoid bearing a primary hydroxy group at C-12. This proposal was confirmed by the HMBC correlations of the oxymethylene protons with C-5, C-8, and C-10 of the spatane tricyclic nucleus ( Figure S29). The NOESY correlations ( Figure  S30) indicated that the relative configuration of compound 3 was identical to that of 1 and 2.
Okaspatol D (4) possessed the molecular formula C 20 H 32 O, determined by HRESIMS. The NMR spectra (Table 2) were related to those of the spatanes discussed above. However, the 1 H NMR spectrum of 4 did not show any signal attributable to protons geminal to an oxygenated function. In fact, the 13 C NMR spectrum of 4 only showed a carbon linked to oxygenated function at δ C 82.2, which was identified as C-10 from its HMBC correlations with H-3, H-5, H-8, H-9, and Me-12 ( Figure S29). On the other hand, compound 4 was concluded to possess a side chain that differed from that of 1 and 3 by the E configuration of the double bond at C-13,C-15, as indicated by the shielding of Me-14 (δ H 1.59/δ C 17.9) [29] and the NOESY correlation of Me-14 with the methylene protons H-16 ( Figure S30). In addition to the new okaspatols A-D (1-4) described above, the known spatanes 5-7 were also isolated. Compounds 5 and 6 have already been found in R. okamurae [14]. Compound 7 has been previously described from R. marginata, although only partial and not assigned NMR data were reported [5]. The fully assigned NMR data of compound 7 (Table S1), as well as the key COSY, HMBC, and NOESY correlations, are provided in Figures S29 and S30.
On the basis of COSY and HMBC correlations (Figure 4a), the two double bonds mentioned above, together with three methyl groups (δ H 1.69 (Me-14, Me-19), δ H 1.65 (Me-20)) and a bis-allylic methylene (δ C 28.1 (C-16)/δ H 2.84, 2.72 (H-16)), were accommodated in a 1,5-dimethylhexa-1,4-dien-1-yl side chain identical to that found in compounds 1, 3, and 4-7. The Z configuration of the double bond at C-13,C-15 was deduced from the NOESY correlation Me-14/H-15. The position of this chain on the cyclopentanecarbaldehyde ring was supported by the HMBC correlations H-6/C-13, H-7/C-14, H-8/C-13, and H-15/C-7. The HMBC correlation of the proton geminal to the acetoxy group (δ H 5.63, H-5) with C-7 and C-8 determined the presence of the acetoxy group also on this ring, at C-5. On the other hand, the hydroxy group was deduced to be located on the ciclopentanone ring, at C-2, from the HMBC correlations of the proton geminal to the hydroxy group (δ H 4.10, H-2) with C-9, the ketone carbonyl at C-10, and the methyl group at C-11. in a 1,5-dimethylhexa-1,4-dien-1-yl side chain identical to that found in com and 4-7. The Z configuration of the double bond at C-13,C-15 was dedu NOESY correlation Me-14/H-15. The position of this chain cyclopentanecarbaldehyde ring was supported by the HMBC correlations 7/C-14, H-8/C-13, and H-15/C-7. The HMBC correlation of the proton ge acetoxy group (δH 5.63, H-5) with C-7 and C-8 determined the presence o group also on this ring, at C-5. On the other hand, the hydroxy group was d located on the ciclopentanone ring, at C-2, from the HMBC correlations o geminal to the hydroxy group (δH 4.10, H-2) with C-9, the ketone carbonyl the methyl group at C-11.   for dilkamural [11] and rugukamurals A-C [15]. Based on biogenetic considerations, the stereochemistry of one ring with respect to the other in rugukamural D (8) is depicted as identical to that established for dilkamural [11].
Together with rugukamural D (8), the four known secospatanes 9-12 were also isolated. Compounds 9-11 are known diterpenoids of R. marginata [5] and herein are described for the first time as metabolites of R. okamurae. Compound 12 was first isolated from R. marginata [5] and later from R. okamurae [13]. Since partial and not assigned NMR data were reported for these four compounds [5], fully assigned 1 H and 13 C NMR spectra of 9-12, together with key COSY, HMBC, and NOESY correlations, are provided in Table S2 and Figures S29 and S30.

Prenylcubebane Diterpenoids 13-15
The molecular formula C 20 H 30 O of okacubol A (13) was determined by HRESIMS. This formula, together with the presence of five methyl signals in the 1 H NMR spectrum at δ H 1.83 (3H), 1.80 (3H), 1.27 (6H), and 1.00 (3H) ( Table 4), suggested that 13 was also a diterpenoid. The NMR spectra showed the signals of a tertiary oxygenated carbon (δ C 71.4) and of three double bonds (δ C 143.3,140.4,137.9,126.2,123.0,118.1), bearing in total four olefinic protons (δ H 6.52, 5.83, 5.67, 5.25). Since the three double bonds accounted for three of the six degrees of unsaturation deduced from the molecular formula, compound 13 must be tricyclic. Two of the double bonds mentioned above, the oxygenated carbon and three methyl groups, were deduced to be located in a 5-hydroxy-1,5-dimethylhexa-1,3-dien-1-yl side chain. Key data were the HMBC correlations of the oxygenated carbon (δ C 71.4, C-18) with two methyl groups (δ H 1.27, Me-19 and Me-20) and with two olefinic protons (δ H 5.67 (H-17) and 6.52 (H-16)), together with the COSY coupling of these with another olefinic proton (δ H 5.83 (H-15)), which showed allylic coupling with the methyl group at δ H 1.80 (Me-14) (Figure 5a). On the other hand, the remaining methine and methylene groups of the molecule were connected following the sequence of COSY couplings spanning from the olefinic methine proton at δ H 5.25 (H-5) to the methylene at δ H 2.08 (H-3a)/1.74 (H-3b). Then, the HMBC correlations H-3/C-1,C-2,C-9, and C-10 were consistent with a five-membered ring, while the correlation H-3/C-8 defined the presence of the fused threemembered ring. The tricyclic framework was completed from the correlations H-3/C-4, H-5/C-10, and Me-12/C-10. Moreover, the HMBC correlation H-7/C-14 indicated that the side chain was connected to C-7. All these data led to conclude that compound Two of the double bonds mentioned above, the oxygenated carbon and three methyl groups, were deduced to be located in a 5-hydroxy-1,5-dimethylhexa-1,3-dien-1-yl side chain. Key data were the HMBC correlations of the oxygenated carbon (δC 71.4, C-18) with two methyl groups (δH 1.27, Me-19 and Me-20) and with two olefinic protons (δH 5.67 (H-17) and 6.52 (H-16)), together with the COSY coupling of these with another olefinic proton (δH 5.83 (H-15)), which showed allylic coupling with the methyl group at δH 1.80 (Me-14) (Figure 5a). On the other hand, the remaining methine and methylene groups of the molecule were connected following the sequence of COSY couplings spanning from the olefinic methine proton at δH 5.25 (H-5) to the methylene at δH 2.08 (H-3a)/1.74 (H-3b). Then, the HMBC correlations H-3/C-1,C-2,C-9, and C-10 were consistent with a five-membered ring, while the correlation H-3/C-8 defined the presence of the fused three-membered ring. The tricyclic framework was completed from the correlations H-3/C-4, H-5/C-10, and Me-12/C-10. Moreover, the HMBC correlation H-7/C-14 indicated that the side chain was connected to C-7. All these data led to conclude that compound The molecular formula C20H30O determined by HRESIMS of okacubol B (14) indicated that it was an isomer of compound 13. The analysis of the NMR spectra (Table 4) showed that 14 possessed the same tricyclic core as 13 but a different side chain. In particular, the hydroxy group was deduced to be located at C-13, and the conjugated double bonds at C-15,C-16, and C-17,C-18. Key data were the HMBC correlations of the methyl group at δH 1.42 (Me-14), both with the oxygenated carbon (δC 78.4, C-13) and with C-7 in the ring ( Figure S29). The 15Z configuration was deduced from the coupling constant J15,16 = 12.1 Hz.
In addition to compounds 13 and 14, the known prenylcubebane 15 was also isolated. This compound was described from R. marginata [5], although only partial NMR data were reported. The full NMR data of compound 15, together with key COSY, HMBC, and NOESY correlations, are shown in Table S3 and Figures S29 and S30.

Prenylkelsoane Diterpenoid 16
The molecular formula C20H32O of okamurol A (16) was determined by HRESIMS, and the IR absorption band at 3384 cm −1 indicated the presence of a hydroxy group. The NMR spectra (Table 5) showed the signals of two double bonds, one 1,1-disubstituted (δC 150.0 (C-13) and δC 110.5 (C-14)/δH 4.94,4.85 (H-14)) and one trisubstituted (δC 125.6 (C- The molecular formula C 20 H 30 O determined by HRESIMS of okacubol B (14) indicated that it was an isomer of compound 13. The analysis of the NMR spectra (Table 4) showed that 14 possessed the same tricyclic core as 13 but a different side chain. In particular, the hydroxy group was deduced to be located at C-13, and the conjugated double bonds at C-15,C-16, and C-17,C-18. Key data were the HMBC correlations of the methyl group at δ H 1.42 (Me-14), both with the oxygenated carbon (δ C 78.4, C-13) and with C-7 in the ring ( Figure S29). The 15Z configuration was deduced from the coupling constant J 15,16 = 12.1 Hz.
In addition to compounds 13 and 14, the known prenylcubebane 15 was also isolated. This compound was described from R. marginata [5], although only partial NMR data were reported. The full NMR data of compound 15, together with key COSY, HMBC, and NOESY correlations, are shown in Table S3 and Figures S29 and S30.

Prenylkelsoane Diterpenoid 16
The molecular formula C 20 H 32 O of okamurol A (16) was determined by HRESIMS, and the IR absorption band at 3384 cm −1 indicated the presence of a hydroxy group. The NMR spectra (Table 5) showed the signals of two double bonds, one 1,1-disubstituted (δ C 150.0 (C-13) and δ C 110.5 (C-14)/δ H 4.94,4.85 (H-14)) and one trisubstituted (δ C 125.6 (C-17)/δ H 5.08 (H-17) and δ C 132.2 (C-18)), which together with two allylic methylenes (δ C 37.7(C-15)/δ H 1.91,1.84 (H-15) and δ C 27.4 (C-16)/δ H 2.09,2.05 (H-16)) and two methyl groups (δ H 1.65 (Me-19), δ H 1.59 (Me-20)) were accommodated in a 6-methylhepta-1,5-dien-2-yl side chain similar to that found in 2. Eliminating these eight carbons and two unsaturations from the molecular formula showed that the remaining twelve carbons of the molecule must be arranged in a tricyclic moiety. In fact, similar to the spatane 4, the NMR data indicated that the tricyclic core of 16 contained four methylenes, four methines, and two fully substituted carbons, one of them linked to the hydroxy group. However, a careful analysis of the COSY and HMBC correlations (Figure 6a) indicated that 16 did not exhibit the fused 5/4/5 tricyclic ring system characteristic of the spatanes, but an alternative 5/5/4 arrangement of kelsoane type [30], featuring a cyclopentane ring and a cyclobutane ring fused to a central cyclopentane ring. unsaturations from the molecular formula showed that the remaining twelve carbons of the molecule must be arranged in a tricyclic moiety. In fact, similar to the spatane 4, the NMR data indicated that the tricyclic core of 16 contained four methylenes, four methines, and two fully substituted carbons, one of them linked to the hydroxy group. However, a careful analysis of the COSY and HMBC correlations (Figure 6a) indicated that 16 did not exhibit the fused 5/4/5 tricyclic ring system characteristic of the spatanes, but an alternative 5/5/4 arrangement of kelsoane type [30], featuring a cyclopentane ring and a cyclobutane ring fused to a central cyclopentane ring.  In particular, the presence of a methyl-substituted five-membered ring was supported by the HMBC correlations of the methine proton at δH 2.59 (H-9), with two methylene carbons (δC 32.2 (C-2) and 34.4 (C-3)) and with the oxygenated carbon (δC 94.6, C- In particular, the presence of a methyl-substituted five-membered ring was supported by the HMBC correlations of the methine proton at δ H 2.59 (H-9), with two methylene carbons (δ C 32.2 (C-2) and 34.4 (C-3)) and with the oxygenated carbon (δ C 94.6, C-10), together with the sequence of COSY couplings of protons δ H 2.59 (H-9)/δ H 2.50 (H-1)/δ H 0.92 Me-11. Key data to define the central five-membered ring were the COSY sequence of couplings δ H 2.59 (H-9)/δ H 2.38 (H-8)/δ H 2.29 (H-7) together with the HMBC correlation of H-7 with C-9 and of Me-12 with C-7 and C-10. Moreover, the location of the side chain linked to this central ring was deduced from the HMBC correlation of H-9 with the olefinic carbon at δ C 150.0 (C-13). The presence of the four-membered ring was deduced from two COSY-coupled methylenes (δ H 2.05, 1.40 (CH 2 -5) and δ H 1.89, 1.52 (CH 2 -6)), which showed HMBC correlations with the oxygenated carbon C-10 and with the carbon C-8 bearing the side chain, respectively. All these data led to defining that compound 16 possessed a carbon skeleton of prenylkelsoane. The relative configuration of compound 16 was determined from the NOESY data. Thus, the correlation H-7/Me-12 indicated the cis fusion between the four-and five-membered rings (Figure 6b). The correlations of H-3a with H-8, Me-12, H-2b, and H-2b and H-8 with Me-11 indicated that H-8, Me-11, and Me-12 were on the same face of the molecule. On the other face were H-9 and the side chain, as supported by the NOESY correlation between H-9 and H-14. Although the absolute configuration has not been determined, compound 16 has been depicted on the basis of biogenetic considerations (see below), with the same configuration as the spatanes and other diterpenoids isolated from this alga.
The prenylkelsoane carbon skeleton exhibited by compound 16 is very unusual. A survey of the literature revealed only one previous account, which describes the isolation of a deoxyanalogue of 16 from the mixture resulting from the incubation of geranylgeranyl diphosphate (GGPP) with a terpene synthase isolated from the marine bacterium Streptomyces xinghaiensis [31]. It is worth noting that in that study, the diterpene spata-13,17-diene was the major compound derived from GGPP, leading to the proposal of the formation of the tricyclic core of spatanes and prenylkelsoanes from a common biosynthetic intermediate, which was supported by isotope labeling experiments [31]. The biosynthesis of spatanes and the prenylkelsoane okamurol A (16) in the alga Rugulopteryx could take place in a similar way, as shown in Figure 7. After the cyclization of GGPP to yield the (E,E)-prenylgermacradienyl cation A, a second cyclization would produce cation B. This could experience two alternative openings of the cyclopropane ring to give C and D, which are precursors of the spatanes and of the prenylkelsoane 16, respectively. 10), together with the sequence of COSY couplings of protons δH 2.59 (H-9)/δH 2.50 (H-1)/δH 0.92 Me-11. Key data to define the central five-membered ring were the COSY sequence of couplings δH 2.59 (H-9)/δH 2.38 (H-8)/δH 2.29 (H-7) together with the HMBC correlation of H-7 with C-9 and of Me-12 with C-7 and C-10. Moreover, the location of the side chain linked to this central ring was deduced from the HMBC correlation of H-9 with the olefinic carbon at δC 150.0 (C-13). The presence of the four-membered ring was deduced from two COSY-coupled methylenes (δH 2.05, 1.40 (CH2-5) and δH 1.89, 1.52 (CH2-6)), which showed HMBC correlations with the oxygenated carbon C-10 and with the carbon C-8 bearing the side chain, respectively. All these data led to defining that compound 16 possessed a carbon skeleton of prenylkelsoane. The relative configuration of compound 16 was determined from the NOESY data. Thus, the correlation H-7/Me-12 indicated the cis fusion between the four-and five-membered rings (Figure 6b). The correlations of H-3a with H-8, Me-12, H-2b, and H-2b and H-8 with Me-11 indicated that H-8, Me-11, and Me-12 were on the same face of the molecule. On the other face were H-9 and the side chain, as supported by the NOESY correlation between H-9 and H-14. Although the absolute configuration has not been determined, compound 16 has been depicted on the basis of biogenetic considerations (see below), with the same configuration as the spatanes and other diterpenoids isolated from this alga.
The prenylkelsoane carbon skeleton exhibited by compound 16 is very unusual. A survey of the literature revealed only one previous account, which describes the isolation of a deoxyanalogue of 16 from the mixture resulting from the incubation of geranylgeranyl diphosphate (GGPP) with a terpene synthase isolated from the marine bacterium Streptomyces xinghaiensis [31]. It is worth noting that in that study, the diterpene spata-13,17-diene was the major compound derived from GGPP, leading to the proposal of the formation of the tricyclic core of spatanes and prenylkelsoanes from a common biosynthetic intermediate, which was supported by isotope labeling experiments [31]. The biosynthesis of spatanes and the prenylkelsoane okamurol A (16) in the alga Rugulopteryx could take place in a similar way, as shown in Figure 7. After the cyclization of GGPP to yield the (E,E)-prenylgermacradienyl cation A, a second cyclization would produce cation B. This could experience two alternative openings of the cyclopropane ring to give C and D, which are precursors of the spatanes and of the prenylkelsoane 16, respectively.

Inhibitory Activity of NO Production
The anti-inflammatory activity of ten of the isolated compounds, including representatives of the four classes of diterpenoids, was assayed in specific immune cell lines, microglial cells Bv.2, and macrophage cells RAW 264.7. In particular, the capability of the spatane diterpenoids 1, 3, 5, and 6, the secospatanes 8, 10, 11, and 12, the prenylcubebane 14, and the prenylkelsoane 16 to inhibit the lipopolysaccharide(LPS)-stimulated production of nitric oxide (NO) in Bv.2 cells and RAW 264.7 cells was tested. Compounds 7, 9,

Inhibitory Activity of NO Production
The anti-inflammatory activity of ten of the isolated compounds, including representatives of the four classes of diterpenoids, was assayed in specific immune cell lines, microglial cells Bv.2, and macrophage cells RAW 264.7. In particular, the capability of the spatane diterpenoids 1, 3, 5, and 6, the secospatanes 8, 10, 11, and 12, the prenylcubebane 14, and the prenylkelsoane 16 to inhibit the lipopolysaccharide(LPS)-stimulated production of nitric oxide (NO) in Bv.2 cells and RAW 264.7 cells was tested. Compounds 7, 9, and 13 were not tested because of solubility issues or low stability, while compounds 2 and 4 were expected to cause similar effects to compounds 5 and 6.
First, in order to select the highest dose of each compound that could be used in the assays without affecting cell viability, the cytotoxicity of each diterpenoid against Bv.2 and RAW 264.7 cells was evaluated (Figures S31 and S32). The spatane diterpenoids 1, 3, 5, 6, and the prenylkelsoane 16 were the less toxic compounds, and in most instances, they did not affect cell viability at concentrations up to 10 µM or even 25 µM for compound 3. On the other hand, the secospatanes 8, 10, 11, and 12 were much more toxic for both microglial and macrophage cells, and at 10 µM caused more than 70% of cell death for Bv.2 and more than 60% of cell death for RAW 264.7. The prenylcubebane 14 did not affect Bv.2 cells viability at concentrations up to 1 µM and RAW 264.7 up to 10 µM. On this basis, the highest doses of the compound used in the NO inhibitory assays with Bv.2 and RAW 264.7 cells were those shown in Table 6 in order to guarantee that decreases of NO levels in the cells were truly due to inhibition of its production and not to cell death. Table 6. Highest concentration (µM) of diterpenoids 1, 3, 5, 6, 8, 10, 11, 12, 14, and 16 that does not affect cellular viability of Bv2 cells and RAW 264.7 cells. For the anti-inflammatory assays, cells were pretreated for 3 h with the compounds at the highest non-cytotoxic doses previously evaluated (Table 6) and then stimulated with LPS (200 ng/mL) to trigger an inflammatory response, which includes the overproduction and release of NO. The production of this inflammatory mediator was determined by measuring the levels in the culture medium of nitrite secreted by the cells, which is one of the major metabolites derived from NO. The treatment of cells with the compounds in the absence of LPS did not cause any induction of inflammatory response. The changes observed in NO production in microglial cells Bv.2 and macrophage cells RAW 264.7 after pretreatment with the compounds and LPS stimulation are shown in Figures 8 and 9, respectively. 5, 6, and the prenylkelsoane 16 were the less toxic compounds, and in most instances, did not affect cell viability at concentrations up to 10 μM or even 25 μM for compoun On the other hand, the secospatanes 8, 10, 11, and 12 were much more toxic for both croglial and macrophage cells, and at 10 μM caused more than 70% of cell death for and more than 60% of cell death for RAW 264.7. The prenylcubebane 14 did not affect cells viability at concentrations up to 1 μM and RAW 264.7 up to 10 μM. On this basis highest doses of the compound used in the NO inhibitory assays with Bv.2 and RAW 2 cells were those shown in Table 6 in order to guarantee that decreases of NO levels in cells were truly due to inhibition of its production and not to cell death.  diterpenoids 1, 3, 5, 6, 8, 10, 11, 12 For the anti-inflammatory assays, cells were pretreated for 3 h with the compou at the highest non-cytotoxic doses previously evaluated ( Table 6) and then stimul with LPS (200 ng/mL) to trigger an inflammatory response, which includes the over duction and release of NO. The production of this inflammatory mediator was determ by measuring the levels in the culture medium of nitrite secreted by the cells, which is of the major metabolites derived from NO. The treatment of cells with the compound the absence of LPS did not cause any induction of inflammatory response. The chan observed in NO production in microglial cells Bv.2 and macrophage cells RAW 264.7 a pretreatment with the compounds and LPS stimulation are shown in Figures 8 and 9 spectively.  and subsequently stimulated with LPS. Upon stimulation of cells with LPS, the nitrites level increased threefold compared to control Bv.2 cells. However, treatment of cells with diterpenoids 1, 3, 5, 6, 11, 12, 14, and 16 significantly inhibited nitrite production, with decreases of 38.8%, 65.0%, 28.1%, 43.0%, 27.6%, 41.1%, 29.1%, and 54.4%, respectively, in relation to LPS-stimulated but non-treated cells. The inhibition caused by okaspatol C (3) and okamurol A (16) was outstanding, since effectively counteracted the effect of LPS stimulation. Upon treatment with these compounds, and in spite of adding the inflammatory stimulus, the level of nitrites remained identical or close to that of basal conditions and similar to the levels observed in cells treated with the classical anti-inflammatory steroid dexamethasone. Both compounds showed dose-dependent inhibitory activity ( Figure 8B). The strongest inhibition was caused by compound 3 at 25 µM, although significant effects were also observed at a concentration of 10 µM, with compound 16 being more active than 3. The spatanes 1 and 6 and the secospatane 12 also caused noticeable inhibitory effects, suppressing the LPS-induced overproduction of nitrites to less than half.
As shown in Figure 9A, the treatment of macrophage cells RAW 264.7 with LPS induced the overproduction of nitrites. The pretreatment of cells with compounds 3, 5, 6, 12, 14, and 16 caused inhibition of the nitrite level by 64.0%, 30.8%, 28.9%, 35.0%, 39.0%, and 43.2%, respectively, in relation to LPS-stimulated but non-treated cells. Similar to that observed with Bv.2 cells, compound 3 completely suppressed the effects of LPS stimulation in RAW 264.7 cells. The inhibition of nitrite production caused by compound 16 was also noticeable, although more moderated, suppressing the LPS-induced overproduction of nitrites to less than half. Both compounds, 3 and 16, showed dose-dependent inhibitory activity ( Figure 9B). The strongest inhibition was observed for 3 at 25 µM, while at 10 µM, compound 16 was more active than 3. The spatane 5, the secospatane 12, and the prenylcubebane 14 were also capable of reducing the overproduction of NO to half. Figure 8A shows the nitrite production measured in cultures of control Bv.2 cells (B bar), in cultures of cells stimulated with LPS (LPS bar), and in cultures of cells treated with the algal diterpenoids at the selected concentration or with dexamethaxone (DX bar) and subsequently stimulated with LPS. Upon stimulation of cells with LPS, the nitrites level increased threefold compared to control Bv.2 cells. However, treatment of cells with diterpenoids 1, 3, 5, 6, 11, 12, 14, and 16 significantly inhibited nitrite production, with decreases of 38.8%, 65.0%, 28.1%, 43.0%, 27.6%, 41.1%, 29.1%, and 54.4%, respectively, in relation to LPS-stimulated but non-treated cells. The inhibition caused by okaspatol C (3) and okamurol A (16) was outstanding, since effectively counteracted the effect of LPS stimulation. Upon treatment with these compounds, and in spite of adding the inflammatory stimulus, the level of nitrites remained identical or close to that of basal conditions and similar to the levels observed in cells treated with the classical anti-inflammatory steroid dexamethasone. Both compounds showed dose-dependent inhibitory activity ( Figure  8B). The strongest inhibition was caused by compound 3 at 25 μM, although significant effects were also observed at a concentration of 10 μM, with compound 16 being more active than 3. The spatanes 1 and 6 and the secospatane 12 also caused noticeable inhibitory effects, suppressing the LPS-induced overproduction of nitrites to less than half.
As shown in Figure 9A, the treatment of macrophage cells RAW 264.7 with LPS induced the overproduction of nitrites. The pretreatment of cells with compounds 3, 5, 6, 12, 14, and 16 caused inhibition of the nitrite level by 64.0%, 30.8%, 28.9%, 35.0%, 39.0%, and 43.2%, respectively, in relation to LPS-stimulated but non-treated cells. Similar to that observed with Bv.2 cells, compound 3 completely suppressed the effects of LPS stimulation in RAW 264.7 cells. The inhibition of nitrite production caused by compound 16 was also noticeable, although more moderated, suppressing the LPS-induced overproduction of nitrites to less than half. Both compounds, 3 and 16, showed dose-dependent inhibitory activity ( Figure 9B). The strongest inhibition was observed for 3 at 25 μM, while at 10 μM, compound 16 was more active than 3. The spatane 5, the secospatane 12, and the prenylcubebane 14 were also capable of reducing the overproduction of NO to half.  Although diterpenoids of the spatane class have been known for a long time, data on their biomedical properties are scarce and mostly focused on the cytotoxicity shown by a few compounds in assays against cancer cell lines [17,18,25,32,33]. This study has shown  1, 1, 10 µM) on NO release in RAW 264.7 cells. RAW 264.7 cells were pretreated for 3 h with the compound at the corresponding concentration or with dexamethasone (DX, 2.5 µM), followed by stimulation with 200 ng/mL LPS for 24 h. Nitrite accumulation in the culture media was measured using the Griess reagent. Results are expressed as fold change relative to the LPS condition and are mean ± SD (n ≥ 3 independent experiments performed in duplicate). Significant differences were determined by two-way ANOVA followed by Bonferroni t-test. * p ≤ 0.05 vs. LPS. # p ≤ 0.05 vs. Basal.
Although diterpenoids of the spatane class have been known for a long time, data on their biomedical properties are scarce and mostly focused on the cytotoxicity shown by a few compounds in assays against cancer cell lines [17,18,25,32,33]. This study has shown the anti-inflammatory properties of various spatane diterpenoids, with okaspatol C (3) outstanding because of its capacity to completely inhibit the overproduction of NO induced upon inflammatory stimulation, both in microglial Bv.2 and in macrophage RAW 264.7 cells. On the other hand, most of the compounds of the secospatane series tested in this study showed weak or no significant inhibitory activity. This outcome strengths observations made in our previous study where another six secospatanes were assayed [15]. Altogether, the results suggest that the 1,5-dicarbonyl system present in secospatanes 8, 10, 11, 12, dilkamural [15], and rugukamural C [15] enhances cytotoxicity, constraining the use of these compounds to very low doses, which may not be effective in NO inhibition. On the other hand, secospatanes, such as ruguloptones A, B, and F, exhibiting a -CH 2 OR group at C-12 instead of the aldehyde function, are not toxic and effective NO inhibitors [15]. NO inhibitory activity was also detected for the prenylcubebane 14 and with more potency for the prenylkelsoane 16, although only a compound of each of these diterpenoid classes has been tested.
During the last three decades, growing evidence has shown that enhanced NO production, due to the expression of inducible nitric oxide synthase (iNOS), plays a key role in the pathophysiology of inflammation [34,35]. Thus, inhibition of iNOS is an important strategy to control the inflammatory processes associated with many pathological conditions. In this regard, structurally diverse natural products, mostly isolated from plants, have been reported to act by inhibiting NO production [35]. In recent years, several metabolites of macroalgae, including phenolic compounds [36][37][38], fucosterol [36], fucoxanthin [39], C 15 acetogenins [40], and terpenoids [40][41][42], have been described to decrease NO levels in LPS-stimulated macrophage RAW 264.7 cells and fucosterol also in microglial cells [43]. Together with our previous report [15], this study has demonstrated the NO inhibitory activity of algal diterpenoids of four structural classes in both immune cells Bv.2 and RAW 264.7 cells, showing the potential of macroalgal terpenoids in the search for lead compounds for new anti-inflammatory agents.