Absolute configuration by vibrational circular dichroism of anti-inflammatory macrolide briarane diterpenoids from the Gorgonian Briareum asbestinum

The four new briarane diterpenoids 2-butyryloxybriarane B-3 (2), 9-acetylbriarenolide S (3), briarenolide W (4), and 12-isobriarenolide P (5), along with briarane B-3 (1) and the five known diterpenes 6–10 were isolated from the gorgonian coral Briareum asbestinum collected from the Mexican Caribbean Sea. The structures were elucidated by 1D and 2D NMR and MS measurements. Since the structure of briarane B-3 (1) was only suggested and published without any spectroscopic support, it was herein confirmed, and the supporting data are now provided. In addition, 1 provided the opportunity to explore the sensitivity of vibrational circular dichroism (VCD) to determine the configuration of a single stereogenic center in the presence of eight other stereogenic centers in a molecule possessing a highly flexible ten-member ring. A single-crystal X-ray diffraction study, in which the Flack and Hooft parameters of 1 were determined, further confirmed that briarane B-3 is (1S,2S,6S,7R,8R,9S,10S,11R,17R)-1. This paper reports for first time the use of VCD in briarane diterpenes and with the presence of chlorine atoms. Biological evaluation of seven isolated compounds evidenced a moderate anti-inflammatory activity for compounds 6 and 9 but it did not show any cytotoxic, antiviral, antibacterial, and topoisomerase inhibitory activity.

www.nature.com/scientificreports/ C-1, between H-14 and C-1 and C-12, between H-10 and C-11, between H 3 -20 and C-10, C-11, and C-12, and between H 3 -15 and C-1, C-2, and C-14 (Fig. 2). The γ-lactone was located on the briarane scaffold by the HMBC correlations between H 3 -18 and C-8 and the γ-lactone C-19 carbonyl signal at δ C 176.4, and between H-17 and C-8 and C-9. In addition, the presence of two acetate groups was evidenced by two methyl singlets at δ H 2.14 (δ C 21.9) and δ H 2.24 (δ C 21.2) along with the carbonyl signals at δ C 170.3 and 169.6. The HMBC correlations between these carbonyl signals and the methine doublets at δ H 4.81 and 5.11, respectively, (Fig. 2), defined the positions of the acetate groups at C-2 and C-9. The NMR data of 1 were similar to those of solenolide E 28 in which C-9 has a hydroxy group.
SciFinder and MarinLit databases searches revealed that the (1S,2R,6S,7R,8R,9R,11R,17R)-1 stereostructure was suggested (no stereochemistry at C-10), based on undisclosed NMR methods, for an isolate named as briarane B-3 29 . The compound was found in the same coral, also collected from the Caribbean Sea, although at several West Indies locations instead from a mainland location. The C-10 stereogenic center was not drawn at the time 29 , but it was recently assumed 30 as (10S) from the lack of a NOE effect of the angular methyl group and H-10, in agreement with a trans ring fusion present in similar briaranes 31 . A (1S,2S,6S,7R,8R,9R,10S,11R,17R) stereochemistry was assumed based on extensive NMR studies including NOESY, J-HMBC, and HSQC-HEC-ADE measurements 30 . Other observed NOE effects that can be considered as secure for 1 are those of the methyl groups at C-1 and C-11 which have an 1,3-diaxial distribution on a cyclohexenone, as well as that seen for H-7 and H-17 which are syn distributed on the γ-lactone, although in this case the observation is not defining the relative stereochemistry of H-7 and H-10. In contrast, a NOE effect observed for H-2 and H-10 has to be interpreted with care since either stereochemistry at C-2 is feasible according to the conformational freedom of the ten-member ring, which is evident after assembly and manipulation of a solid Dreiding stereomodel from Büchi (Flawil, Switzerland), as was done to understand the uncommon conformational behavior of the diterpenoid icetexone 32 . In addition, there is no useful NOE effect observation to define the C-9 stereogenic center.
The stereochemistry of 1 could be expected as that of solenolide E 28 , although, as mentioned above, there appeared to be uncertainty for the configuration of the C-2 and C-9 stereogenic centers. Since the absolute configuration (AC) of 1 could be verified at the same time using an independent methodology, vibrational circular dichroism (VCD) was selected for this purpose. This methodology has shown to be a powerful tool 33 for the AC determination of many types of natural products 34 , but not yet for briaranes. It was also successful for the determination of one stereogenic center in the presence of nine other stereogenic centers, as occurred for the epoxidation reaction study of the pentacyclic triterpenoid lupenone 35 .
We begin with IR and VCD spectra calculation for the (1S,2R,6S,7R,8R,9R,10S,11R,17R) diastereoisomer by constructing a molecular model in the Spartan 04 software. This model was submitted to conformational  36 without any conformational restriction, in particular since we are dealing with a molecule possessing a ten-member ring. This search provided 25 conformers in a 9.93 kcal/mol energy window for which the single point (SP) energy was calculated at the DFT B3LYP/6-31G(d) level of theory using the same software. The five conformers found in the initial 5 kcal/mol energy gap, were optimized at the DFT B3LYP/DGDZVP level of theory using the Gaussian 09 software to render three conformers contributing each with more than 1% to the total conformational distribution. These minimum energy structures were verified for the absence of imaginary frequencies, they are shown in Figure S49 (Supplementary Information), and their thermochemical data are summarized in Table S3 in Supplementary Information. These three conformers were used for the IR and VCD frequency calculations   www.nature.com/scientificreports/ at the same level of theory, also in the Gaussian 09 suit. Conformer population weighting was done according to the ΔG = − RT ln K equation to generate the respective Boltzmann averaged IR and VCD spectra. The IR and VCD spectra of the (1S,2R,6S,7R,8R,9R,10S,11R,17R) diastereoisomer were compared to the experimental IR and VCD spectra of 1 ( Figure S50 in Supplementary Information) using the statistic correlation of the CompareVOA program 37 . The confidence level data are summarized in Table S2 in Supplementary Information, where the optimal anharmonicity factor (anH) was 0.972. The IR similarity (S IR ) was relatively poor (89.3) and the VCD similarity for the assumed stereochemistry (S E ) was only 24.7, while that for the other enantiomer (S −E ) was 45.3, providing a very poor Enantiomer Similarity Index (ESI) of − 220.6, which is obtained as the S E − S −E difference. The 51% confidence level (C) for the (1S,2R,6S,7R,8R,9R,10S,11R,17R) AC of 1 clearly showed the relative stereochemistry was incorrectly assumed.
Since a NOE effect for H-10 and H-2 was seen, but no effect was evident for H-9, it was next assumed this stereogenic center might be inverted. Thus, a similar calculation protocol was undertaken for the (1S,2R,6S,7R,8R,9S,10S,11R,17R) diastereoisomer. The MMFF search provided 11 conformers, in a 9.06 kcal/ mol gap, which were submitted to SP calculations. The six conformers in the initial 5 kcal/mol energy gap, were DFT B3LYP/DGDZVP optimized to render the four conformers shown in Figure S51 (in Supplementary Information) contributing each with more than 1% to the conformational distribution. The conformers, whose thermochemical parameters are summarized in Table S2 in Supplementary Information, were used for the IR and VCD frequency calculations and also to generate the IR and VCD spectra. A spectrum contrasting procedure using the CompareVOA software provided the data given in Table S3 while the plots are shown in Figure S52 (in Supplementary Information).
A good data improvement revealed the (9S) configuration is indeed evident since S IR improved to 93.3, S E almost duplicated to 48.9, S −E severely decreased to 25.6, and C increased to 62%. The above results suggested the C-2 configuration should also be inverted. MMFF calculation of the (1S,2S,6S,7R,8R,9S,10S,11R,17R) diastereoisomer provided 52 conformers in a 9.98 energy gap which after SP calculations left seven conformers in the initial 5 kcal/mol gap. Energy optimizations, as above, rendered the four conformers shown in Fig. 3 which reveal the high conformational freedom of the ten-member ring, while the corresponding thermochemical data are summarized in Table S2 in Supplementary Information.
These four conformers were used to calculate the IR and VCD spectra of the (1S,2S,6S,7R,8R,9S,10S,11R,17R)diastereoisomer. CompareVOA contrasting of the calculated and experimental spectra revealed that the two acetates of 1 are beta oriented and have the (S) absolute configuration since S IR climbed to an excellent 97.3 value, S IR now is 78.0, and the confidence level (C) is 100%. The corresponding spectra are shown in Fig. 4. www.nature.com/scientificreports/ To complete the study for determining the C-2 and C-9 stereochemistry, the fourth possibility, the (1S,2S,6S,7R,8R,9R,10S,11R,17R)-diastereoisomer was also calculated as above. The 41 conformers obtained in a 9.77 kcal/mol gap, after MMFF searches, were reduced to nine in a 5 kcal/mol gap after SP calculations.
The thermochemical parameters of the three final contributing conformers are given in Table S2 in Supplementary Information and the CompareVOA data are summarized in Table S3, while the conformers are shown in Figure S53 (Supplementary Information) and the spectra comparison in Figure S54 (Supplementary Information). Close inspection of Table S3 shows that the (2S,9R) and (2R,9S) data sets provide similar S IR and C values since in either case one stereogenic center is incorrectly assumed. In addition, ESI is negative in the (2S,9R) case since the S -E value is larger than the S E value, indicating that the C-2 and C-9 stereogenic centers highly influence the overall VCD spectroscopic behavior. It also follows from the VCD study of the four C-2 and C-9 diatereoisomers that the stereostructure originally suggested for (1S,2S,6S,7R,8R,9S,10S,11R,17R)-1 is indeed correct and it is evident that VCD is a very sensitive methodology to establish the absolute configuration (AC) of a single stereogenic center, located on a flexible macrocycle, in the presence of eight other stereogenic centers. To our knowledge, this is the first report of the use of VCD in briarane diterpenes, and also with a natural compound possessing a chlorine atom.
The absolute configuration of 1 was independently confirmed in the present study by single-crystal X-ray diffraction (XRD). This method has been used successfully a long time ago for briaranes from B. asbestinum when they possess a halogen atom, like briarein A 8 . In that case the chlorine atom scattering anomalous dispersion was used to measure differences in Bijvoet intensities, allowing to establish the AC of the compound. In the present www.nature.com/scientificreports/ case, two independent measurements were done. One XRD study, performed at 100 K, allowed to refine the data to a convergence value R = 3.5% as detailed in the Experimental Section. In the second XRD study the data were acquired at room temperature for the complete diffraction sphere, thereby allowing determining the AC by means of the the Flack 38 and Hooft 39 parameters. The R value was 4.8%, while the Olex2 software 40 provided the Flack and Hooft parameters for (1S,2S,6S,7R,8R,9S,10S,11R,17R)-1 as x = − 0.010 (13) and y = 0.007(10), respectively, which for the inverted structure were x = 1.003 (13) and y = 1.012(10), respectively. Figure 5 depicts the AC on an ORTEP drawing for 1.
The absolute determination of 1 by VCD and XRD is in agreement with a very recent 1 H residual chemical shift anisotropy measurement 41 , a powerful contemporary methodology for the study of molecular structures 42 .
Structure elucidation of 2-10. Compound 2 was also isolated as an amorphous white powder. Its molecular formula of C 26 H 35 ClO 8 was deduced from its ( +)-HRESIMS. The IR spectrum and the 1 H-and 13 C-NMR data (Tables 1, 2) were very similar to those of 1, but they displayed that one of the acetate groups in 1 had been replaced by a butyrate moiety in 2. This was suggested from its 13 C-NMR spectrum that shows two ester-type carbonyl carbons (δ C 170.4 and 172.6) and from its 1 H-NMR spectrum showing only one acetate methyl signal at δ H 2.15 and an additional spin system consisting of a methyl triplet signal at δ H 1.03 correlated by 1 H-1 H COSY to a multiplet at 1.77 (m), and this in turn to another multiplet centered at 2.45 (m). Two additional sp 3 methylenes observed in the 13 C-NMR of 2 and the increase of 28 uma of its molecular weight in relation to that of 1 confirmed the substitution of one of the two acetates in 1 for a butyrate group in 2. The relative configuration of 2 was displayed to be the same as that of 1 by the identical NOE correlations ( Figure S55, Supporting Information) observed in its NOESY experiment, the similar proton-proton coupling constants in its 1 H-NMR and by comparison of the NMR data of 2 with those of 1 (Tables 1, 2). The similar optical rotation of 2 in relation to that of 1 suggests that they share the same absolute configuration. So, we named compound 2 as 2-butyryloxybriarane B-3.
The C 24 H 31 ClO 8 molecular composition of 3, isolated as an amorphous white powder, followed from its HRESIMS which showed the [M + Na] + ion at m/z 505.1596 (calcd for C 24 H 31 ClO 8 Na, 505.1605). This composition is the same as that of 1 and therefore both compounds seem to be isomers. Comparison of the 1 H NMR spectra of 1 and 3 (Table 1) (Table 1) and the C-16 signal in 3, at δ C 51.2, is shifted to δ C 67.5 in 4 ( Table 2). The relative configuration of 4 at C-1, C-2, C-7, C-8, C-9, C-10, C-11, and C-17 is the same than in 1-3 as evidenced by similar correlations observed in its NOESY plot ( Figure S57 in Supplementary Information). We have named 4 as briarenolide W.
Briarane 5 was also isolated as an amorphous white powder. The isotopes clusters at m/z 479/481 in a 3:1 ratios corresponding to [M + Na] + ion peak revealed that 5 also had a chlorine atom. The HRESIMS of 5 at m/z 479.1437 (calculated 479.1449), indicated the C 22 H 29 ClO 8 molecular composition with an IHD of eight. Comparison of the IR, 1 H, and 13 C NMR data of 5 showed they were very similar to those of briarenolide P 31 , the main differences being found for the six-member ring, and remarkably for the C-20 methyl signal. The strong NOE effect of H 3 -15 and H 3 -20 indicated that they are on the same face of the molecule. The shift of C-20 from δ C 9.1 in briarenolide P to δ C 13.3 in 5 suggested a smaller γ-gauche effect due to an 1,3-diaxial interaction between these two methyl groups in 5 as compared to that occurring in briarenolide P 31 . This suggests a conformational change of the six-member ring, probably due to the different configuration of the hydroxy group at C-12 (see 13 C chemical shifts in red and blue colors in Fig. 6).
Thus, the relative stereochemistry of the hydroxy and epoxy groups at C-12, C-13, and C-14 in 5 followed from evaluation of their 1 H NMR coupling constants, the heteronuclear carbon-proton coupling constants in a HSQC-HECADE experiment, and its NOESY experiment, using our reported 43 approach for the evaluation of 3β,7-dihydroxy-5,6-epoxycholestanes. The 3 J H-12/H-13 = 5.9 Hz and 3 J C-13/H-14 = 1 Hz values in 5 implied the cis disposition of the hydroxy and epoxy groups. Besides, the strong NOESY cross peak observed between Me-20 and H-12 places both protons on the same face of the ring. The almost identical proton and carbon chemical shifts and proton-proton coupling constants observed in the NMR spectra of 5 and briarenolide P (see black colored carbon chemical shifts in Fig. 6), suggests that they have the same relative configuration at all other stereogenic centers. It thus followed that 5 is the C-12 epimer of briarenolide P, and therefore we named it as 12-isobriarenolide P.
Five known diterpenoids were identified as brianthein X (6)  Anti-inflammatory activity. In order to assure the lack of toxicity of the molecules before more detailed biological studies, the seven diterpenes (compounds 1, 4, 6-10) were initially assed for their potential toxicity against human non-cancer cells, namely keratinocytes (HaCaT). At the highest concentration tested (100 µM) no toxicity was detected ( Figure S48 in Supplementary Information).
Posteriorly, the diterpenes were evaluated for their anti-inflammatory capacity, as assessed for their ability to inhibit or lower the activation of the nuclear transcription factor NF-κB (Fig. 7A). The experimental model used consisted of human THP-1 monocytes stably transfected with a NF-κB-inducible Luc reporter construct (Fig. 7B). Cells were then differentiated into macrophages by incubation with phorbol-myristate acetate (PMA) for 24 h, after which the M1 phenotype was induced by exposure to lipopolysaccharide (LPS). Preliminary studies evaluated the impact of the diterpenes in the viability of these cells, in order to ensure that only non-toxic www.nature.com/scientificreports/ concentrations were used. As shown in Figure S48, no toxicity was detected at 100 µM. All molecules were evaluated for their impact in NF-κB signaling at two concentrations, 25 and 100 µM. As it can be seen in Fig. 7A, compounds 6 and 9 significantly reduced NF-κB activation at the highest concentration tested. Considering that the NF-κB transcription factor is considered a master switch in controlling the inflammatory cascade, the capacity of compounds 6 and 9 to lower its activation status is of interest, as these molecules can be further studied as leads for anti-inflammatory drugs. To the best of our knowledge, briarane diterpenes have been described before as anti-inflammatory agents in mouse cells 51 ours being the first study to describe such effect in human macrophages.
Seeing these results, we were interested in investigating which downstream genes could be modulated by 6 and 9 to exert their anti-inflammatory activity. In Fig. 7C we present the normalized expression of mRNA for COX-2, IL-6, IL-1β and TNF-α. As shown in the results, 6 and 9 significantly decrease the expression of COX-2 and IL-6, with 9 displaying the same activity towards TNF-α.
All the isolated compounds were also submitted to additional biological evaluations including cytotoxic, topoisomerase I and antibacterial activities. They did not show significant cytotoxic activity against the human tumor cell lines (MDA-MB-231 (breast), HT-29 (colon), NSCLC A549 (lung) and PSN1 (pancreas)); topoisomerase inhibitory activity against the human topoisomerase I, nor antibacterial activity against Acinetobacter baumannii, Pseudomonas aeruginosa, Klebsiella pneumoniae, and Staphylococcus aureus.
The insecticidal activity of compound 6 was previously reported 44,52 . The cytotoxic activity and neuroactivity of compound 9 was previously reported 48,53 .

Conclusions
In summary, the four new briarane diterpenoids, 2-butyryloxybriarane B-3 (2), 9-acetylbriarenolide S (3), briarenolide W (4), and 12-isobriarenolide P (5), together with previously suggested briarane B-3 (1), the briantheins 6-8, asbestinin (9), and diterpenoid 10 were found in the gorgonian Briareum asbestinum collected from the Mexican Caribbean Sea. The originally suggested structure and stereochemistry of 1 was established by a combination of NMR measurements and VCD spectroscopy in combination with DFT calculations for first time for this type of compounds. In fact, due to the high conformational flexibility of the ten-member ring of 1, which www.nature.com/scientificreports/ difficults the relative configuration determination of several stereogenic centers by means of NMR methods, allows to suggest that VCD might be the method of choice for the determination of the relative configuration of specific stereogenic centers on a ten-member ring when crystals suitable for X-ray diffraction studies could not be gained. VCD has the additional advantage that it provides the absolute configuration of the studied molecule. Several biological tests using the isolated metabolites revealed a mild anti-inflamatory activity.

Methods
General experimental procedures. Optical rotations were measured on a JASCO DIP-1000 polarimeter, with a Na (589 nm) lamp and filter. IR spectra were measured on a FTIR Bruker Vector 22 spectrometer. 1 H, 13 C, and 2D NMR spectra were recorded on a Bruker Avance 500 spectrometer at 500 and 125 MHz, respectively, using CDCl 3 . Low resolution electrospray mass spectrometry (LRESIMS) and high-resolution electrospray mass spectrometry (HRESIMS) experiments were performed on the Applied Biosystems QSTAR Elite system. The chemical shifts were given in δ (ppm) and coupling constants in Hz. HPLC separations were performed on the Agilent 1100 liquid chromatography system equipped with a solvent degasser, quaternary pump, and diode array detector (Agilent Technologies, Waldbronn, Germany) using a semipreparative normal phase column Novapak, silica 6 µm 60 Å, 300 × 7.8 mm (Waters were exhaustively extracted with a mixture of CH 3 OH-CH 2 Cl 2 (1:1, 3 × 1.5L) at 25 °C for 24 h each extraction, and the extracts were combined and concentrated under vacuum. Following the modified Kupchan methodology, the dark green crude residue (20.0 g) was first partitioned between CH 2 Cl 2 /H 2 O (1:1 v/v). Then, the organic phase was concentrated under reduced pressure and partitioned between 10% aqueous MeOH (400 mL) and hexane (2 × 400 mL). The H 2 O content (% v/v) of the methanolic fraction was adjusted to 50% aqueous MeOH and this mixture was extracted with CH 2 Cl 2 (100 mL). The CH 2 Cl 2 -soluble portion was subjected to flash column chromatography using silica gel and a gradient of n-hexane/EtOAc (0-100%) to obtain 14 fractions (C1-C14). Fraction C9-C11 was separated by NP-HPLC using an isocratic mixture of 80:20 hexane-acetone as the mobile phase, to obtain compounds asbestinin-10 (9, 10 mg), 2 (1 mg), lactone-14 (10, 4 mg), 3 (2 mg), 1 (25 mg) and brianthein Z (8, 10 mg). Crystallization of fraction C10H8 (n-hexane/EtOAc, 4:1) furnished brianthein Y (7) (8 mg). Fractions C12 and C13 were purified by NP-HPLC using an isocratic mixture of 60:40 hexane-acetone as the mobile phase, to obtain compounds brianthein X (6, 11 mg), 4 (2 mg) and 5 (2 mg). Vibrational circular dichroism studies. The data were acquired on a BioTools dualPEM ChiralIR FT spectrophotometer using a solution of 5.0 mg of 1 in 150 μL of 100% D atom CDCl 3 . The solution was placed in a cell having BaF 2 windows and a path-length of 100 μm. The data were measured at a resolution of 4 cm −1 for 6 h and the base-line was provided by subtracting the spectrum of the solvent acquired under identical experimental conditions. The stability of 1 during the data acquisition procedure was verified by 1 H NMR spectroscopy immediately before and after the VCD measurement. The in silico constructed molecular models of the four diastereoisomers of 1 at C-2 and C-9 were subjected to Monte Carlo search protocols in a 10 kcal/mol energy window using Merck Molecular Force Field (MMFF94) as implemented in the Spartan'04 program without any conformational restriction. These searches provided 25 conformers for the (2R,9R) diastereoisomer, 11 for the (2R,9S) diastereoisomer, 52 for the (2S,9S) diastereoisomer, and 41 for the (2S,9R) diastereoisomer in 9.93, 9.06, 9.98, and 9.77 kcal/mol energy windows, respectively. These 129 conformational models were subjected to single point energy calculation using DFT at the B3LYP/6-31G(d) level of theory in the same software suit. These procedures provided five, six, seven, and nine conformers, in the same order respectively, found in the initial 5 kcal/mol. These 28 conformers were subjected to geometry optimization and energy calculation by DFT using the B3LYP/DGDZVP level of theory as implemented in the Gaussian 09 program. The three, five, four, and three minimized structures, in the same order respectively, contributing individually with more than 1% to the total conformational distribution of each diastereoisomer, provided the thermochemical parameters given in Table S2 in Electronic Supporting Information and were used to calculate the IR and VCD frequencies at 298 K and 1 atm. All minimum energy structures were verified for the absence of imaginary frequencies and their relative free energies were employed to calculate their Boltzmann population. The Boltzmann-weighted IR and VCD spectra were calculated considering Lorentzian bands with half-widths of 6 cm −1 . Molecular visualization was accomplished using the GaussView 6.0 program. Geometry optimization and vibrational calculations required some 65 h of CPU time per conformer when using a PC operated at 3.5 GHz and 4 Gb RAM.
MTT viability assay. Monocytes were seeded at a density of 6.0 × 10 4 cells/well. PMA (50 nM) was added as a differentiation agent to obtain macrophages. After 24 h, this medium was discarded and replaced with fresh PMA-free medium for another 24 h period, after which the differentiated M1-macrophages were incubated with the compounds of interest for 24 h. After this period the wells were aspirated, and the medium replaced with MTT at 0.5 mg/mL and incubated for 2 h. At the end of this period, the solution was discarded and the formazan crystals in the well dissolved in 200 µL of a DMSO:isopropanol solution (3:1). The absorbance at 560 nm was read in a Thermo Scientific Multiskan GO microplate reader. For HaCaT cells, the density used was 1.5 × 10 4 cells/well. NF-κB activation assay. THP-1 Lucia NF-κB monocytes were seeded on 96-well plates and differentiated into macrophages as described above for non-transfected THP-1 monocytes and incubated with the selected compounds. After 2 h, LPS from E. coli was added to each well at a final concentration of 1 µg/mL. 22  www.nature.com/scientificreports/ µL of supernatant were collected from each well and transferred to a white 96-well plate. Then, as indicated by the supplier, 50 µL of QUANTI-Luc assay solution was added to each well, the plate was shaken, and luminescence was immediately read in a Cytation 3 (BioTek) microplate reader.
RNA extraction, quantification, integrity, and conversion. THP-1 cells were seeded at a density of 4.8 × 10 5 cells/well in 12-well plates and differentiated into macrophages as described above and treated for 22 h with LPS (1 µg/mL), with or without pre-incubation with compounds 6 and 9 for 2 h. Afterwards, the supernatant was removed, and the cells were disrupted in 500 µL of PureZOL reagent. Then, the samples were transferred to a RNase-free tube and 100 µL of chloroform were added. The mixture was shaken vigorously for 15 s. After 5 min incubation at room temperature, the samples were centrifuged at 12,000 × g for 15 min at 4 °C. Following centrifugation, the aqueous phase containing the RNA was immediately transferred to a new RNase-free tube and 250 µL of isopropyl alcohol were added, and mixture was incubated at room temperature for 5 min. Afterwards, the tubes were centrifuged at 12,000 × g for 10 min at 4 °C, the RNA appearing as a white pellet on the side and bottom of the tube. Supernatant was carefully discarded, and the RNA pellet was washed with 1 mL of 75% of ethanol. After vortexing, the mixture was centrifuged at 7500 × g for 5 min at 4 °C and the supernatant was carefully discarded. Then, the RNA pellet was air-dried for about 5 min and reconstituted in 25 µL of PCR grade water. Subsequently, the RNA was quantified in a Qubit 4 fluorometer (Invitrogen by Thermo Fisher Scientific; Waltham, MA, USA), using the Qubit RNA HS assay kit. The RNA quality and integrity were then evaluated using the Qubit RNA IQ assay kit. In order to obtain the complementary DNA, 1 µg of RNA was mixed with 4 µL qScript cDNA SuperMix in a 20-µL reaction. The reverse-transcribed reaction involved three steps: 5 min at 25 °C, 30 min at 42 °C, and 5 min at 85 °C 54 .
q-PCR analysis. q-PCR analysis were conducted on multiple genes, namely COX-2, TNF-α, IL-6, and IL-1β (Table S9 in ESI). GAPDH was used as reference gene (Table S9 in ESI). The primers were designed using the Primer-BLAST tool (NCBI, Bethesda, MD, USA) and synthesized by Thermo Fisher (Waltham, MA, USA), as listed in Table S9 in the Supplementary Information. Real-time qPCR was performed with 2 ng of cDNA using KAPA SYBR FAST qPCR Kit Master Mix (2X) Universal similarly to what we have described before 55 . The thermal cycling conditions were as follows: 3 min at 95 °C, followed by 40 cycles of denaturation at 95 °C for 3 s, specific annealing temperatures for each gene ( Table S9 In the Supplementary Information) for 20 s, and extension at 72 °C for 20 s. The fluorescence signal was detected at the end of each cycle. The results were analyzed with qPCRsoft 4.0 supplied with the equipment qTOWER3 G (Analytik Jena AG, Germany), and a melting curve was used to confirm the specificity of the products. Transcript abundances of the target genes were normalized to the expression of GAPDH (reference gene). Samples were run in duplicate in each PCR assay. Normalized expression values were calculated following the mathematical model proposed by Pfaffl using the formula: 2 − ΔΔCt 56 At least three independent experiments were performed.