Myrtinols A–F: New Anti-Inflammatory Peltogynoid Flavonoid Derivatives from the Leaves of Australian Indigenous Plant Backhousia myrtifolia

Our in-house ethnopharmacological knowledge directed our anti-inflammatory investigation into the leaves of Backhousia mytifolia. Bioassay guided isolation of the Australian indigenous plant Backhousia myrtifolia led to the isolation of six new rare peltogynoid derivatives named myrtinols A–F (1–6) along with three known compounds 4-O-methylcedrusin (7), 7-O-methylcedrusin (8) and 8-demethylsideroxylin (9). The chemical structures of all the compounds were elucidated by detailed spectroscopic data analysis, and absolute configuration was established using X-ray crystallography analysis. All compounds were evaluated for their anti-inflammatory activity by assessing the inhibition of nitric oxide (NO) production and tumor necrosis factor- α (TNF-α) in lipopolysaccharide (LPS) and interferon (IFN)-γ activated RAW 264.7 macrophages. A structure activity relationship was also established between compounds (1–6), noting promising anti-inflammatory potential by compounds 5 and 9 with an IC50 value of 8.51 ± 0.47 and 8.30 ± 0.96 µg/mL for NO inhibition and 17.21 ± 0.22 and 46.79 ± 5.87 µg/mL for TNF-α inhibition, respectively.


Introduction
Inflammation is the body's natural defense response to various harmful stimuli including pathogens, heat, toxic chemicals, and injuries [1]. During the initial stage of trauma or infection, body initiates various cellular and molecular events which include the secretion of many proinflammatory cytokines and chemokines such as interleukin-6 (IL-6), interleukin-1β (IL-1β), tumor necrosis factor-α (TNF-α) as well as reactive oxygen species (ROS) and nitric oxide (NO) to restore tissue homeostasis and resolve acute inflammation [2,3]. The proinflammatory cytokines and/or bacterial lipopolysaccharides (LPS) can activate inducible nitric oxide synthase (iNOS) to produce continuously high concentrations of NO, which can then further induce injury at inflammatory sites [4]. Therefore, suppressing the overproduction of NO appears to be a promising strategy for the control of inflammatory diseases.
Backhousia myrtifolia (Myrtaceae) which is commonly referred to as carrol, neverbreak, iron wood, grey myrtle or cinnamon myrtle is a small rainforest tree species that grows in subtropical rainforests regions of Eastern Australia [5][6][7]. It was first discovered and subsequently used by the Indigenous communities of Australia, where the leaves were used to treat colic babies. B. myrtifolia is also known to harbor oils that have cinnamon-like aroma displaying both anti-bacterial and anti-fungal properties [8].
Our ongoing search to discover new anti-inflammatory molecules led us to explore the leaves of B. myrtifolia resulting in the isolation and identification of six new peltogynoid type were used to treat colic babies. B. myrtifolia is also known to harbor oils that have cinnamon-like aroma displaying both anti-bacterial and anti-fungal properties [8].

Structural Elucidation
In addition to this, we were successful in generating a crystal for compound 2, which was then subjected to X-ray crystallography and the data obtained assisted in confirmation of the structure (Figure 4, Table S1) and helped resolve the absolute stereochemistry for the chiral centers C-2 and C-3 as R and S. From the NMR data alone, a large coupling of J H-2/H-3 (9.5 Hz) also suggested a trans confirmation. In addition to this, we were successful in generating a crystal for compound 2, which was then subjected to X-ray crystallography and the data obtained assisted in confirmation of the structure (Figure 4, Table S1) and helped resolve the absolute stereochemistry for the chiral centers C-2 and C-3 as R and S. From the NMR data alone, a large coupling of JH-2/H-3 (9.5 Hz) also suggested a trans confirmation.     A close comparison of the spectroscopic data of 3 with 2 confirmed that myrtinol C possessed the same ring system as myrtinol B, with the only exceptions being the positional change of the hydroxy and methoxy groups present in the ring A and the introduction of a hydroxy functionality C-4 (δ H 5.01, δ C 70.6). This above structural changes were supported by HMBC correlation from H-4 (δ H 5.01) to C-3 (δ C 78.3), C-9 (δ C 154.0) and C-10 (δ C 103.8), from H-8 (δ H 6.15) to C-6 (δ C 106.8), C-7 (δ C 160.0), C-9 and C-10, from H-6 (δ H 1.98) to C-5, C-6, C-7, and from H-7 (δ H 3.77) to C-7 (Tables 1, 2 5) showed the closest similarity to its analogue myrtinol A (1). The only difference in the NMR data and spectra was the detection of an additional methoxy resonance, which was confirmed to be attached at C-7 based on HMBC correlations of 7-OMe (δ H 3.80) to C-7 (δ C 158.8) and H-8 (δ H 6.36) to C-7.
Myrtinol F (6) was obtained as a green sticky mass. The molecular formula was revealed as C 21  Comparison of the spectroscopic data of 2 and 6 showed that myrtinol F possessed the same ring system as myrtinol B, with the only change being the replacement of hydroxy substituent at C-7 by a methoxy group, which was supported by HMBC correlation from H-8 (δH 6.36) C-7(δ C 157.8), and 7-OMe (δ H 3.80) to C-7.
We also take into account that the level of purity of myrtinol F was not 100%, with the possibility of a terpene like impurity present in this fraction ( Figure S35). Low yields and the level of difficulty experienced separating this two-compound mixture prevented us from obtaining an absolutely pure sample of myrtinol F. However, the HRMS data and the clear key NMR resonances attributed to myrtinol F allowed for its complete structural assignment.
The absolute stereochemistry for the chiral centers C-2 and C-3 for the remaining myrtinol analogues was assigned as R and S, respectively, based on the large coupling constant between J H-2/H-3 which was suggestive of a trans configuration along with the consideration of a likely biosynthetic relationship to myrtinol B (2) (Tables 1 and 2) [9,10].

Anti-Inflammatory Activity
All compounds were assessed for their anti-inflammatory activity by evaluating the inhibition of NO production and TNF-α in LPS plus interferon (IFN)-γ activated RAW 264.7 macrophages. All compounds were also evaluated for their cytotoxicity using the Alamar blue assay (Table 3).

Discussion
As shown in Table 3, the varying anti-inflammatory activity depended mostly on the functional groups attached to rings A, B and C. Based on the slight structural variations around the tetracyclic backbone of myrtinols A-F (1-6), we have attempted to evaluate the observed structural-activity relationship (SAR) among them. A SAR was observed between compounds 1 and 5 (both having a 1,3-benzodioxole moiety attached to ring B) which showed promising inhibition of NO production and TNF-α production, with a IC 50 values of 11.47 ± 0.14, 24.54 ± 0.28 and 8.51 ± 0.47, 17.21 ± 0.22 µg/mL, respectively. In comparison, compounds 2 and 6 showed NO production inhibition with an IC 50 value of 16.25 ± 0.77 and 12.62 ± 0.26 µg/mL, and TNF α production inhibition with an IC 50 value 52.35 ± 7.47, and 30.55 ± 5.01 µg/mL, respectively, suggesting that the presence of 1,3-benzodioxole moiety (ring E) and methoxy group at C-5 and C-7 might play an important role in their anti-inflammatory activity. Interestingly, when comparing the SAR between compounds 3 and 4, the only change between them being the introduction of a hydroxyl group at C-4 in 3, this rendered the molecule inactive compared to the baseline activity observed for compound 4 where the NO inhibition was determined to be 29.31 ± 10.95 µg/mL.
We are mindful of the fact that the experimental NO inhibition values obtained for myrtinols may not be a true reflection of their anti-inflammatory activity profile mainly due to the fact that they mostly have a low therapeutic index associated with them. This suggests that future investigations on assessing their cytotoxic activity need to be performed in order to re-evaluate their potential as cytotoxic agents.
Among the known compounds, Compound 9 displayed interesting activity with an IC 50 value 8.30 ± 0.96 µg/mL and IC 50 value of 46.79 ± 5.87 µg/mL, whereas 7 and 8 did not show good anti-inflammatory activity (Table 3).

General Experimental Procedures
UV spectra were recorded on an Agilent Carry UV-Vis Multicell Peltier spectrometer. NMR spectroscopic data were recorded on a Bruker Avance 600 MHz spectrometer (Bruker Biospin GmbH, Germany). HRMS (High Resolution Mass Spectrometry) was carried out using a Waters SYNAPT G2-Si mass spectrometer operating in the positive and negative ESI mode.

Plant Material
The leaves of B. myrtifolia were collected from the Australian Botanic Garden at Mount Annan (NSW, Australia). A voucher specimen (2005-0104) has been deposited at the Australian Botanic Gardens, at Mount Annan, NSW, Australia.
Fr.11 was re-purified by SB-C3 semipreparative column, (250 × 9.4 mm) using a gradient system of 50-60% MeCN/H 2 O with a flow rate of 2mL/min (with a constant 0.01% FA modifier) over 30 min to afford compounds 3 (1.4 mg, t R 22.4 min), and 4 (2.0 mg, t R 23.1 min). Fr.16 was purified using SB-Phenyl semipreparative column, (250 × 9.4 mm) with a gradient system of 10-80% MeCN/H 2 O over 15 mins followed by a change from 80-100% MeCN/H 2 O over 9 mins with a constant flow rate of 2 mL, and then held at 100% MeCN for 2 mins and equilibrated back to 10% MeCN/ H 2 O in 1 min and maintained at this gradient for an additional 2 min, to afford compound 5 (1.9 mg, t R 21.8 min). Fr.17 was repurified using a semi-preparative Agilent C 18 column (5 µm, 9.4

X-ray Crystallographic Analysis of 2
Crystals were obtained by slow cooled evaporation from MeOH: DCM (1:1) solution; suitable crystals were selected for X-ray crystallographic analysis using an MX1 beamline at the Australian Synchrotron, using silicon double crystal monochromated radiation (λ = 0.71073 Å) at 100 K [25]. The XDS software package [26] was used on site for data integration, processing, and scaling. SADABS [27] was used to apply an empirical absorption correction. Shelxt [28] was applied to solve the structure by the intrinsic phasing method, and a suite of SHELX programs [28,29] were used for refinement, via the Olex2 graphical interface [30]. Crystallographic data of 2 (CCDC number: 2236594) was deposited at the Cambridge Crystallographic Data Centre. Additional crystallographic information is available in the Supporting Information (Table S1).

Maintenance of RAW 264.7 Macrophages
Cells were grown in 75 cm 2 flasks on Dulbecco's Modified Eagle Medium (DMEM) containing 10% fetal bovine serum (FBS) that was supplemented with penicillin (100 U/mL), streptomycin (100 µg/mL) and L-glutamine (2 mM). The cell line was maintained in 5% CO 2 at 37 • C, with media being replaced every 3-4 days. Once cells had grown to confluence in the culture flask, they were harvested using a rubber policeman, as opposed to using trypsin, which can remove membrane-bound receptors.

Pro-Inflammatory Activation of Cells
RAW 264.7 cells (1 × 10 6 cells/mL) were seeded in 96 well plates (Corning ® Costar ® , Sigma, Sydney, Australia) overnight until confluency. When the cells were confluent, each compound was serially diluted from a starting concentration of 100 µg mL −1 to construct a dose response curve (i.e., 100, 50, 25, 12.5, 6.25, and 3.13 µg mL −1 ) and co-incubated with cells for 1 hr prior to the addition of 1 µg mL −1 LPS and 10 U mL −1 (1 unit = 0.1 ng/mL) IFN-γ. After activation, the cells were incubated for another 24 hrs at 37 • C. The supernatant was then collected for NO, and TNF-α assays. The cells were subjected to cell viability measurement using the Alamar Blue assay. Non-activated cells (exposed to media only) were used as negative control and activated cells were positive control.

Determination of Nitrite by the Griess Assay
Nitric oxide was determined by the Griess reagent as described in previous studies [31]. Griess reagent was freshly made up of equal volumes of 1% sulfanilamide in 5% phosphoric acid and 0.1% N-1-naphthylethylenediamine dihydrochloride in Milli-Q water. From each well, 50 µL of supernatant was transferred to a fresh 96-well plate and mixed with 50 µL of Griess reagent. The production of nitrite as an indicator of NO was measured at 540 nm in a POLARstar Omega microplate reader (BMG Labtech, Mornington, Australia).

Determination of TNF-α by ELISA
The stored supernatants were analyzed for TNF-α synthesis using a commercial ELISA kit (Peprotech, Brisbane, Australia) according to the manufacturer's instructions. The absorbance was measured at 410 nm [32]. The concentrations of TNF-α in the experimental samples were extrapolated from a standard curve.

Determination of Cell Viability by the Alamar Blue Assay
After various treatments and the stimulation by LPS and IFN-γ overnight, 100 µL of Alamar Blue solution [10% Alamar Blue (resazurin) in DMEM media] was added to cells and incubated at 37 • C for 2 hrs. The fluorescence intensity was measured with excitation at 530 nm and emission at 590 nm using a microplate reader. The results were expressed as a percentage of the intensity to that of control cells (non-activated cells).

Statistical Analysis
Data analysis was carried using GraphPad Prism 9.3.1. Calculations were performed using MS-Excel version 16.61.1. IC 50 values were obtained by using the sigmoidal doseresponse function in GraphPad Prism. The results were expressed as mean ± standard deviation (SD).

Conclusions
In conclusion, we have isolated and characterized six new rare peltogynoid flavonoids from the leaves of Backhousia myrtifolia. Myrtinols exhibited promising anti-inflammatory activity; however, their low therapeutic index warrants further cytotoxic evaluation as part of a future investigation.