All animal experiments were conducted after obtaining the approval of the Seoul National University Institute of Laboratory Animal Resources and Use Committee. Ten weeks old (247.5 ± 54.16 g), were kept in wire cages, in a room with proper, controlled conditions of temperature (21˚C - 23˚C) and humidity (65% - 70%) with 12 h light/dark cycle. The animals were fed with milled pellets, and had free access to water during the experiment. Lipopolysaccharide, derived from Escherichia coli (055:B5) and diluted in a sterile saline solution, was used for induction of periodontitis. One microliter of LPS solution (10 µg/µL) was slowly injected into interdental papilla between the first and second right maxillary molar [23] , two times for fifteen days, using a Hamilton microsyringe. Rats were randomly divided into five experimental groups with six animals in each: (I) injected with saline, treated with vehicle (distilled water); (II) injected with saline, treated with Antifect; (III) injected with LPS, treated with vehicle; (IV) injected with LPS, treated with Antifect; (V) injected with LPS, treated with ASA. Groups I and II received the same volume (1 µL) of sterile saline injection in the same location of rat maxilla. Groups II and IV received Antifect dissolved in water which was administered two times daily by oral gavage (200 mg/kg body weight). Compounds were administered daily during 15 days by oral intubation beginning 24 h after the start of local PBS and LPS injections in the protocol for induction of experimental periodontal disease. Animals were euthanized by cervical dislocation under inhalation anesthesia and the hemi-maxillae were carefully dissected. Gingival soft tissue (approximately 1.5 mm in the frontal plane × 3 mm in the sagittal plane) adjacent to the upper first molars were carefully dissected from 5 samples from each experimental condition, immediately snap-frozen in liquid nitrogen and subsequently stored at −80C until the moment of total protein extraction. The other 5 samples from each experimental condition were subjected to 10% buffered formalin fixation and subsequent histological processing for paraffin embedding (as described below) and used in histological and immunohistochemical analysis.

Fresh and dried roots of woad (Isatis tinctoria L.) were provided in July 2018 by Dr. Ki Hyun Ryoo, Seoul, Republic of Korea, kept for medicinal use. Extracts for screening were prepared as follows: Dried roots were ground (<0.7 mm). An aliquot (100 g) was extracted at room temperature with 80% of 4:6 v/v ratio of CH₂Cl₂ and EtOH (1000 ml) in and ultrasonic bath for 24 hours. Supercritical fluid extraction (SFE) was carried out with a Dionex 703 with 703M Modifier Module (Salt Lake City, USA). Removal of residual solvent was in a rotary vacuum evaporator at 40˚C. The average yield of extraction was 23.17% ± 7.34%.

All hemi-maxillae (with and without soft tissues) were scanned on a microcomputer tomograph (Skyscan 1176, SkyScan Aartselaar, Belgium) using 18 μm slices. Digital radiographic images of each sample were reconstructed into a three dimensional model and a standardized gray scale value was set to distinguish mineralized from non-mineralized tissues. These three dimensional images were re-oriented on the sagittal, coronal and transversal planes in a standardized manner using anatomical landmarks on molar teeth. A standardized region of interest of 9.72 mm³ was defined, including the first molar, the anterior half of the second molar and extending medially (towards the center of the palate) approximately 1 mm from the most palatal aspect of the crown of the first molar. All image reconstruction, reorientation and analysis of bone volume/total volume (BV/TV fraction) in the region of interest was performed using the software package of the scanning equipment software.

The material, molars with the surrounding gingival tissues, was rinsed in saline and fixed in 10% formaldehyde solution for up to 7 days and then subjected to a process of demineralization, due to the presence of bone elements and teeth, for 15 days in a 18% solution of disodium ethylenediaminetetraacetate (EDTA). After the completion of decalcination, the material was cut into thin parts in the mesial-distal direction and treated by conventional histological procedures to paraffin molds. The final preparations were 4 µm thick stained with hematoxylin and eosin (H&E).

Osteoclasts were identified by their morphology, location and positivity for the expression of tartrate-resistant acid phosphatase (TRAP), assessed by immunohistochemistry. Sections mounted on silanized slides were deparaffinized and rehydrated in a decreasing concentration of ethanol. Endogenous peroxidase was blocked with 3% hydrogen peroxide for 30 min. To inhibit non-specific interactions of the primary antibody, the sections were incubated with a 5% skim milk blocking solution. Slides were washed in distilled water and incubated overnight at 4˚C with goat polyclonal antibody for TRAP (cat# sc-30833, Santa Cruz Biotechnology) diluted 1:100 in PBS containing 2% BSA. Controls included omission of the primary antibody and also incubation with the same concentration of irrelevant IgG raised in goat. After washings in distilled water, the sections were incubated with biotinylated anti-goat antibodies for 30 min, washed, incubated with streptavidin conjugated to HRP and finally with the DAB substrate, according to the protocol of the supplier of the visualization system. Subsequently, the sections were counterstained with Mayeŕs Hematoxylin and coverslips mounted. Osteoclasts were identified as large, multinucleated (3 or more visible nuclei) and TRAP-positive cells, located in the vicinity of the alveolar bone, from the apical region of the palatal root of the first molar to a 400 μm distance from the palatal alveolar bone crest towards the midline (center of the palate). This analysis was performed in 4 semi-serial sections (spaced 100 μm in the antero-posterior plane) from each specimen (20 sections for each experimental condition).

Detection of active (cleaved) caspase-3 in the cells of the gingival tissues was performed by immunohistochemistry. The same general protocol described for detection of TRAP was used, with the inclusion of an antigen retrieval step, performed in citrate buffer (pH 6.0) at 95˚C for 20 min. After blocking of endogenous peroxidase and non-specific interactions of the primary antibody, the sections were washed in distilled water and incubated overnight at 4C with 1:50 dilution (in PBS containing 2% BSA) of rabbit polyclonal antibodies against rat active caspase-3 (cat# ab2302, Abcam). A total of 15 semi-serial sections (3 from each specimen from 5 different animals in each experimental condition) were assessed and the number of positively-stained cells counted in the area of interest (described in stereometric analysis) by a trained examiner who was unaware of the experimental condition of the sections.

The levels of the factors of inflammation, TNF-α, IL-1β, and IL-6, were determined using supernatants of 5% homogenate made up of fresh gingival tissues with deionized water. The measurement of cytokines was performed by enzyme-linked immunosorbent assay (ELISA) kits specific for rats in accordance with the instructions provided by the manufacturer (Invitrogen, Life Technologies, USA). Briefly, 50 µL of sample was added to respective wells of a plate followed by 50 µL of a standard diluent buffer, and 50 µL of incubation buffer for IL-6 assay. After incubation and washing of the wells 100 µL of appropriate biotin conjugate was added, then the plates were left to incubate again followed by washing. Next stage included adding of 100 µL and incubation with streptavidin-HRP working solution. After this step and washing, the plates were incubated with 100 µL of stabilized chromogen followed by final adding of 100 µL of stop solution. Absorbance of each well was read at 450 nm on ELISA microplate reader (Multiskan Ascent No354, Thermo Labsystems, Finland). A value of the sample concentration was calculated from a standard curve and twice multiplied. The results are presented as pg of the cytokines in mg of fresh gingival tissue.

Data obtained from each experiment were analyzed using GraphPad Prism 5.0 (GraphPad Software Inc., San Diego, CA, USA). The objective of the analysis was to compare the results of each outcome according to the different tested compounds. The Shapiro-Wilk test was used to verify that the data had a normal distribution. Pairwise comparisons were performed by unpaired Student t-tests with Welch’s correction for unequal variances and by one-way analysis of variance (ANOVA) followed by Tukey test for post-hoc pairwise comparisons. We adopted 5% as significant level (p < 0.05).

Morphological changes in HE preparations are not visible in the Antifect treated group compare to the control. Figure 1(a) displays a cross section of rat molar with preserved structure of the tooth pulp tissue and periodontal ligament with a slight increase in the number of neutrophils. There are no perceptible changes in the alveolar bone. The histological characteristics of the HE slides of the experimental group I and II correspond completely to healthy tissue properties (Figure 1(b) and Figure 1(c)). Compared to the control animals (I and II group) the histopathology of the animals subjected to periodontitis receiving the vehicle (III group) displays notable changes both in soft tissue and hard tissue. The changes are characterized by the expansion of blood vessels in the periodontal tissue and gingival increase in the number of neutrophils and lymphocytes. Inflammation is followed by resorption of alveolar bone with the visible osteoclasts on the surface of bone and increased bone marrow space, strong deepening of the gingival sulcus and its separation from the cement teeth in the direction of the tooth root. The presence of red blood cells out of blood vessels indicating the bleeding (Figure 1(d)) is also evident. In the experimental groups IV and V (Figure 1(d) and Figure 1(e)), which were subjected to Antifect and ASA, it can be observed much smaller number of inflammatory cells in regard to the third group of rats. In periodontal and gingival tissue a larger number of fibroblasts persist compared to other groups of experimental animals, while the gingival sulcus is in the normal position.

Microcomputer tomography analysis indicates that attenuated inflammatory bone resorption in antifect treatment group (Figure 2). Group I and II had no

effect, and specimens from the LPS non treated animals were not scanned because previous Figure 1 shows that this vehicle does not affect the extent of inflammatory bone resorption in this model. The main purpose of this study was to compare the effects of Antifect and vehicle, ASA on alveolar bone resorption associated with the LPS-induced model of periodontal disease in an experiment performed concomitantly, took a μCT images and compare each group. With same histological analysis, alveolar bone level and bony defected area significantly increase in group III. However, with 15 days of compound treatment, increase of alveolar bone level and decrease of defected area also observed in group IV and V (Figure 2(a)). Long-term ASA administration reported recovery of inflammatory symptom of periodontitis patient [24] , but gastric-intestinal irritation limits its long-term treatment. Antifect treatment group (IV) consistently showed decrease of alveolar bone loss similar with ASA treatment group (V) (Figure 2(b)).

Of primary importance, the number of TRAP-positive multinucleated cells, identified by immunohistochemistry, was significantly decreased only in the group IV even not in the group V, which supports the interpretation that the bone-sparing effects of this compound observed in the μCT analysis are due, at least in part, to the inhibition of osteoclastogenesis in vivo. ASA treatment had no effect on the number of osteoclasts present in the periodontal and gingival tissue (Figure 3).

Increased apoptosis in inflammation may generate a loop of sustained activation of the immune response, as dying cells release damage-associated molecular patterns (DAMPs). Inflammation induced by LPS injections increased apoptosis in the periodontal microenvironment, as indicated by significant increases of active caspase-3-stained cells in the gingival tissues. This increase in apoptosis was statistically significant in the alveolar bone crest (osteocytes) in all LPS-treated group. And its distribution focused in the bony defected area and alveolar crest area which was most activated inflammatory area. Administration of Antifect, but not of ASA treatment group, potently and significantly reduced the number of apoptotic cells in the gingival tissues and also of osteocytes in the alveolar bone crest, both in the presence and absence of LPS-induced inflammation (Figure 4).

The treatment with Antifect and ASA (IV and V group) significantly decreased the level of TNF-α, IL-6 and IL-1β in the gingival tissue with LPS-induced periodontitis when compared to the untreated group (III). There was no statistically significant difference of the levels of cytokines between groups which did not receive LPS (I and II). ASA treated group (V group) significantly reduced the levels of IL-1β, but not IL-6 and TNF-α, in comparison to the group that treated antifect (IV) (Figure 5).