Diterpenoid Lactones with Anti-Inflammatory Effects from the Aerial Parts of Andrographis paniculata

Andrographis paniculata (AP) has been widely used in China for centuries to treat various diseases, and especially to treat inflammation. Diterpenoid lactones are the main anti-inflammatory components of AP. However, systematic chemical composition and biological activities, as well as key pharmacophores, of these diterpenoid lactones from AP have not yet been clearly understood. In this study, 17 diterpenoid lactones, including 2 new compounds, were identified by spectroscopic methods, and most of them attenuated the generation of TNF-α and IL-6 in LPS-induced RAW 274.7 cells examined by ELISA. Pharmacophores of diterpenoid lactones responsible for the anti-inflammatory activities were revealed based on the quantitative structure-activity relationship (QSAR) models. Moreover, new compounds (AP-1 and AP-4) exerted anti-inflammatory activity in LPS microinjection-induced zebrafish, which might be correlated with the inhibition of the translocation of NF-κB p65 from cytoplasm to nucleus. Our study provides guidelines for future structure modification and rational drug design of diterpenoid lactones with anti-inflammatory properties in medical chemistry.

To enrich the diversity of natural diterpenoid lactones with better bioactivities, 17 diterpenoid lactones, including 2 new compounds, were isolated and identified from AP and the anti-inflammatory properties of these compounds were systemically evaluated in lipopolysaccharide (LPS)-stimulated RAW 264.7 macrophages. Thereby, we determined the key pharmacophores of diterpenoid lactones and developed a quantitative structure-activity relationship (QSAR) model for illustrating the relationship between the chemical structures and anti-inflammatory effects. Moreover, new compounds (AP-1 and AP-4) inhibited the translocation of NF-κB p65 from cytoplasm to nucleus and attenuated inflammation in LPS-microinjection zebrafish. This research provides useful information for future structure modification and rational design of diterpenoid lactones as potent anti-inflammatory agents.
Pro-inflammatory cytokines interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α) play important roles in the inflammatory response [31,32]. LPS can stimulate the secretion of IL-6 and TNF-α in RAW 264.7 macrophage cells to induce inflammation [33]. To investigate the anti-inflammatory effects of isolated diterpenoid lactones, the levels of IL-6 and TNF-α in LPS-stimulated RAW 264.7 cells were determined. As shown in Table 3, most of compounds could significantly reduce the release of IL-6 and TNF-α. Moreover, AP-6 and AP-8 exerted the best activities, which were similar to that of dexamethasone (Dex, the positive control), indicating their potential as anti-inflammatory agents.

D-QSAR Analysis
Based on the anti-inflammatory effects and the chemical structures of 19 diterpenoid lactones, there is the possibility of developing an alignment rule for the superposition of the diverse derivatives to conduct the 3D-QSAR study by comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) methods. Compound AP-6, with the greatest potential anti-inflammatory effect, was selected as the template for the alignment of all compounds. Figure 4A shows the common skeleton of 19 compounds, and the aligned molecules are shown in Figure 4B. The actual and predicted pI values of the training set and test set are listed in Table 4. The CoMFA and CoMSIA models gave good predictive correlation coefficients (r 2 pred ) of 0.689 and 0.707, respectively, clearly suggesting that the models were reliable and capable of predicting the activity of diterpenoid lactones.  The contour maps of CoMFA steric and electrostatic fields are depicted in Figure 5 (group 1: IL-6 group) and Figure 6 (group 2: TNF-α group), respectively. As shown in Figure 5A, the huge green contour near the R 2 position suggested that a bulky group at this site would be beneficial for the activities. A small yellow contour near the R 3 position indicated that a minor substituent at this site would be favorable. In Figure 5B, the huge blue contour above the R 2 position indicated that an electron-donating group at this site would be favored. One blue contour close to the point of the R 3 site also suggested that an electron-donating group at this position would be helpful for the activity. Figure 6A shows the contour maps of sterically favored (green) and disfavored (yellow) regions, which are similar to those in Figure 5A. In addition, the contour map in Figure 6B is also similar to that in Figure 5B. Additionally, a blue contour around the R 1 position revealed that an electron-donating group at the distal part of R 1 position was favorable. In addition, a small red contour can be observed around the R 3 site, showing the substituent with electron-withdrawing was favored at this point.  The CoMSIA steric, electrostatic, hydrogen bond donor and hydrophobic contours maps are demonstrated in Figure 7 (group 1) and Figure 8 (group 2). The CoMSIA steric and electrostatic contour maps in Figure 7A,B and Figure 8A,B are similar to the CoMFA contour maps shown in Figures 5 and 6, respectively.  For the CoMSIA hydrogen bond donor field contour maps shown in Figure 7C, the cyan contours represent regions where hydrogen bond donor substituents were favorable for anti-inflammatory activity, whereas the purple contours were unfavorable. A huge purple contour along the R 1 site indicates that the hydrogen bond acceptor field at this position was important for the activities. A small purple contour beside the R 3 side chain shows that a hydrogen bond acceptor at this site would be favored. A cyan contour near the R 2 position reveals that a hydrogen bond donor substituent at this position would increase the activity. The hydrogen bond donor fields of group 2 ( Figure 8C) were basically similar to those of group 1 ( Figure 7C).
The contour plots of the hydrophobic field are shown in Figures 7D and 8D. The contour plots of the hydrophobic fields in group 1 are shown in Figure 7D, and these were in accordance with the results of group 2 ( Figure 8D). In Figure 8D, there were large yellow regions near the substituents R 1 and R 3 . This data indicates that the increased substituents in the regions would be of benefit to the biological activity. There is a huge yellow contour around the R 2 position, demonstrating that hydrophobic groups play a critical role in this region.
To sum up, the structure-activity relationship revealed by 3D-QSAR is illustrated in Figure 9. In detail, the hydrophobic and hydrogen bond acceptor groups at the R 1 position are favorable; the bulky, electron-donating, hydrogen bond donner and hydrophobic groups at the R 2 position, and the minor, electron-withdrawing hydrophobic groups at the R 3 position, would increase the activity.

New Diterpenoid Lactones Inhibit the Activation of NF-Kb in LPS-Induced RAW 264.7 Cells
The NF-κB signaling pathway is mainly responsible for the process of inflammatory and other diseases [34,35]. NF-κB p65, a critical regulatory transcription factor, separates from the inhibitory subunit when stimulation occurs and translocates into the nucleus, triggering the transcription of multiple inflammatory genes [36]. To research whether NF-κB participates in the anti-inflammatory process of new diterpenoid lactones (AP-1 and AP-4), we tested the nuclear translocation of NF-κB p65. The results showed that AP-1 and AP-4 prevented the translocation of NF-κB p65 from cytoplasm to nucleus in LPS-stimulated RAW 264.7 macrophages (Figure 10), which indicated that AP-1 and AP-4 exert potent anti-inflammatory effects through the inhibition of NF-κB.

Anti-Inflammatory Effects of New Diterpenoid Lactones in Zebrafish
Zebrafish (Danio rerio) has gained tremendous attention as an animal model for assessing the anti-inflammatory effect due to their physiological similarity to mammals, small size, availability in large quantities, transparent body, and relatively low-cost maintenance [37,38]. LPS can induce systemic inflammation in zebrafish [39]. Considering that new diterpenoid lactones, AP-1 and AP-4 exhibited in vitro anti-inflammatory effects, we further evaluated in vivo anti-inflammatory activities by using an LPS-induced zebrafish model.
The survival rates showed that microinjection of LPS caused 100% mortality within 72 h, whereas AP-1 and AP-4 attenuated the mortality of zebrafish ( Figure 11A). Infiltration of inflammatory cells is pivotal in the progression of inflammation [40]. Hematoxylin and eosin (H&E) staining results showed that inflammatory cell infiltration was clearly observed in the model group. However, the histopathological changes were ameliorated after AP-1 and AP-4 treatment ( Figure 11B). Moreover, neutrophils significantly migrated and were recruited to the inflammatory site in the model group, while AP-1 and AP-4 reduced the number of neutrophils around the yolks, which is similar to Dex, as shown in Figure 12.

General Experimental Procedures
The experimental procedures were conducted as described previously [30].

Extraction and Isolation
The extraction and isolation procedures were performed as previously described [30]. The dried extract of AP (840 g) was suspended in H 2 O and successively partitioned with petroleum ether (PE), ethyl acetate (EtOAc) and n-butyl alcohol (n-BuOH). The extract was concentrated under vacuum to obtain PE fraction (30 g), EtOAc fraction (400 g), n-BuOH fraction (140 g) and water fraction, respectively. The EtOAc fraction (400 g) was subjected to silica gel column chromatography (CHCl 3 -MeOH, 100:0 to 0:100, v/v). Based on TLC profiles, similar fractions were combined to give eight fractions (Fr. 1 to Fr. 8). Fr. 2 was separated on a silica gel column, ODS column and preparative HPLC to yield AP-5 (8 mg), AP-9 (300 mg) and AP-16 (300 mg). Fr. 3 (50 g) was subjected to a silica gel column with gradient conditions (CHCl 3 -MeOH, 100:0 to 0:100, v/v) to produce seven subfractions, which were further purified on a silica gel column, ODS column, preparative HPLC and recrystallized to give AP-1 (

Cell Culture
The murine macrophage cell line RAW 264.7 was purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Cells were cultured in DMEM supplemented with 10% (v/v) fetal bovine serum (Invitrogen) and 1% (v/v) penicillin/streptomycin (Gibco) at 37 • C in a humidified atmosphere with 5% CO 2 .

Cell Viability Assay
The cell viability was examined by MTT assay. RAW 264.7 cells were seeded in a 96-well plate at a density of 5 × 10 3 /well. Then, cells were treated with various concentrations of diterpenoid lactones for 24 h. Subsequently, 30 µL of MTT (5 mg/mL) was added into each well and incubated for another 4 h at 37 • C. The purple formazan crystals in each well were dissolved in 100 µL of DMSO and the absorbance was recorded at 570 nm by a microplate reader (Thermo Fisher Scientific, Waltham, MA, USA).

Enzyme-Linked Immunosorbent Assay (ELISA)
The levels of IL-6 and TNF-α in cell supernatants were quantified with ELISA kits. Firstly, the samples and Cytokine standard were diluted in a range of concentrations, and 100 µL/well was placed to the plate. After ninety-minute incubation at 37 • C, plates were washed with four times washing buffer, and Streptavidin-HRP was added to the wells at 100× dilution and incubated for another 30 min, and then washed again. Finally, TMB and Stop solution were added to measure the absorbance at 450 nm.

Confocal Microscopy
RAW 264.7 cells were cultured on a glass dish at a density of 1 × 10 5 for 24 h. After being preincubated with AP-1 (10 µM) or AP-4 (10 µM) for 2 h, the following steps were carried out as previously described [41].

D-QSAR Modeling
To predict the potential anti-inflammatory effects of diterpenoid lactones, CoMFA and CoMSIA were performed. In group 1 (IL-6 group), 19 compounds were split randomly into training (15 compounds) and test sets (4 compounds), respectively. In addition, in group 2 (TNF-α group), 13 molecules were taken for the construction of training set, and the remaining 6 molecules were used to test the proposed CoMFA and CoMSIA models. The anti-inflammatory constants were expressed in pI values (pI = −log[the generation of pro-inflammatory cytokines after chemical treatment at the concentration of 10 µM]), which were applied as the dependent variables in CoMFA and CoMSIA models. In addition, the following experiments were performed as previously described [42].

Zebrafish Husbandry and Embryo Collection
The maintenance of adult transgenic zebrafish line Tg (MPO: GFP) was under standard conditions (Westerfield, 2000). Zebrafish were maintained under a 14-h light/10-h dark cycle at constant temperature (28 ± 0.5 • C) in an independent flow-through system (pH 7.2-7.6, salinity 0.03-0.04%). In addition, the zebrafish were fed brine shrimps three times a day.
Adult zebrafish were moved into a breeding tank until sufficient numbers of embryos were laid at the bottom of the tank. The embryos were washed and moved to clean Petri dishes filled with egg water containing 0.002% methylene blue (Sigma-Aldrich, St Louis, MO, USA) as a fungicide. Finally, these embryos were cultured in a warm oven (28.5 • C) for subsequent experiments.

LPS-Induced Inflammation Model and Chemical Treatments
Larvae were anesthetized at dpf in egg water containing 0.02% Tricaine (Sigma-Aldrich, St. Louis, MO, USA) and aligned in clear petri dishes coated with 1.5% agarose. Subsequently, the yolks were injected with 2 nL of LPS (2 mg/mL) to construct the lethal endotoxin-infected model [39] by using Cell Microinjector (PM1000, MicroData Instrument, Inc. S. Plainfield, NJ, USA). PBS served as the negative control. Then these larvae were washed with fresh egg water and were divided randomly into a 24-well plate (n = 20/well). The control group and the model group were treated with 2 mL of fresh egg water, and five drug groups were treated with Dex (5 µg/mL, Tianxin, Guangzhou, positive control), AP-1 (5 µM), AP-1 (10 µM), AP-4 (5 µM) and AP-4 (10 µM), respectively. For the survival experiment, egg water was changed and mortality (defined as lack of a discernible heart beat) was monitored daily. For the observation of neutrophils, the larvae were imaged at 12 h post LPS-microinjection using a fluorescence microscope (Olympus, MVX10, Tokyo, Japan).

Histopathological Examination of Zebrafish Larvae
Twelve hours after LPS-microinjection, the zebrafish were fixed immediately with 4% (w/v) paraformaldehyde overnight at room temperature, dehydrated in graded ethanol, embedded with paraffin (Leica, Wetzlar, Germany), and 4-µm-thick sections were cut. After being deparaffinized and stained with H&E (Yuanye Biotech, Shanghai, China), each section was observed under an IX 53 light microscope (Olympus, Tokyo, Japan).

Statistical Analysis
All the data are expressed as mean ± standard error (SE), and statistical analyses were performed using the software GraphPad Prism version 5.0 (GraphPad Prism Software, San Diego, CA, USA). Differences were statistically analyzed by Student's t-test or ANOVA with Tukey's test. p < 0.05 were considered statistically significant.

Conclusions
In conclusion, we isolated and identified 17 diterpenoid lactones from AP including 2 new compounds, most of the 19 diterpenoid lactones (including 2 published diterpenoid lactones) showed potent anti-inflammatory activities by reducing the secretion of IL-6 and TNF-α in LPS-stimulated RAW 264.7 macrophages. Furthermore, we discovered the key pharmacophores of diterpenoid lactones and developed a QSAR model to illustrate the significant relationship between the chemical structures and anti-inflammatory effects. Critically, the 2 new compounds exhibited significant anti-inflammatory activities in vitro and in vivo which might be correlated with the inhibition of the translocation of NF-κB p65 from cytoplasm to nucleus. Our study not only enriches the diversity of diterpenoid lactones in AP, but also provides guides for the further structural modification and rational drug design of diterpenoid lactones with anti-inflammatory activities in medical chemistry.
Author Contributions: W.Z. and J.L. conceived and designed the experiments; L.G., Y.Z., L.D., X.W. and F.L. performed the experiments; L.G., Y.Z., P.S. and J.Y. drafted the manuscript; C.F., L.Y. and W.Y. gave some suggestions on the experiment. All authors read and approved the manuscript.