Baicalin Downregulates Porphyromonas gingivalis Lipopolysaccharide-Upregulated IL-6 and IL-8 Expression in Human Oral Keratinocytes by Negative Regulation of TLR Signaling

Periodontal (gum) disease is one of the main global oral health burdens and severe periodontal disease (periodontitis) is a leading cause of tooth loss in adults globally. It also increases the risk of cardiovascular disease and diabetes mellitus. Porphyromonas gingivalis lipopolysaccharide (LPS) is a key virulent attribute that significantly contributes to periodontal pathogenesis. Baicalin is a flavonoid from Scutellaria radix, an herb commonly used in traditional Chinese medicine for treating inflammatory diseases. The present study examined the modulatory effect of baicalin on P. gingivalis LPS-induced expression of IL-6 and IL-8 in human oral keratinocytes (HOKs). Cells were pre-treated with baicalin (0–80 µM) for 24 h, and subsequently treated with P. gingivalis LPS at 10 µg/ml with or without baicalin for 3 h. IL-6 and IL-8 transcripts and proteins were detected by real-time polymerase chain reaction and enzyme-linked immunosorbent assay, respectively. The expression of nuclear factor-κB (NF-κB), p38 mitogen-activated protein kinase (MAPK) and c-Jun N-terminal kinase (JNK) proteins was analyzed by western blot. A panel of genes related to toll-like receptor (TLR) signaling was examined by PCR array. We found that baicalin significantly downregulated P. gingivalis LPS-stimulated expression of IL-6 and IL-8, and inhibited P. gingivalis LPS-activated NF-κB, p38 MAPK and JNK. Furthermore, baicalin markedly downregulated P. gingivalis LPS-induced expression of genes associated with TLR signaling. In conclusion, the present study shows that baicalin may significantly downregulate P. gingivalis LPS-upregulated expression of IL-6 and IL-8 in HOKs via negative regulation of TLR signaling.


Introduction
Periodontal disease is one of the main global oral health burdens and severe periodontal disease (periodontitis) is a major cause of tooth loss in adults globally [1]. Emerging evidence shows that it also increases the risk of some life-threating diseases like cardiovascular disease and diabetes mellitus [2][3][4]. Periodontitis is characterized by bacteria-induced, uncontrolled inflammatory destruction of tooth-supporting tissues and alveolar bone in susceptible individuals [5]. Porphyromonas gingivalis is a major periodontal pathogen and its lipopolysaccharide (LPS) is one of the key virulent attributes that significantly contributes to periodontal pathogenesis [6,7]. It can stimulate the host to produce a variety of pro-inflammatory cytokines like IL-6 and IL-8, thereby involving in the initiation and progression of periodontal disease [8][9][10].
Recently, host modulatory therapy (HMT) has been proposed as a promising adjunct to conventional periodontal treatment [30,31]. Some examples of HMT in treatment of periodontitis include subantimicrobial dose of doxycycline, lipoxins and resolvin E1 [32][33][34]. Scutellariae radix is an herb that has been used to treat inflammatory diseases in traditional Chinese medicine (TCM) since ancient times [35]. Baicalin is a flavonoid isolated from Scutellaria radix and it can suppress IL-8-induced metalloproteinase-8 (MMP-8) expression in human neutrophils [36]. In periodontal research, it has recently been shown that baicalin enables to inhibit the transcription of receptor activator of NF-kB ligand (RANKL) in human periodontal ligament cells, and reduces the loss of bone and collagens in rat models of periodontitis [37,38]. Furthermore, baicalin may inhibit IL-1b-induced MMP-1 expression and stimulate collagen-I production in human periodontal ligament cells [39].
In the present study, we found that baicalin significantly downregulated P. gingivalis LPS-upregulated expression of IL-6 and IL-8. Baicalin also inhibited P. gingivalis LPS-induced activation of NF-kB, p38 MAPK and JNK proteins, and markedly downregulated P. gingivalis LPS-induced expression of genes associated with TLR signaling, such as chemokine (C-C motif) ligand 2 (CCL2), granulocyte colony-stimulating factor (G-CSF or CSF3) and CXCL10.

Cell Culture
HOKs isolated from normal human oral mucosa (Sciencell, CA, USA) were cultured according to the manufacturer's instructions. Prior to cell culture, culture vessels were coated with poly-L-lysine (Sigma, MO, USA) at 2 mg/cm 2 at 37uC for 1 h. Cells were seeded at 5000 cells/cm 2 with the oral keratinocyte medium (Sciencell). The incubation condition was set at 37uC with an atmosphere of 5% CO 2 and 95% air. The medium was changed every two days for the first four days and daily thereafter until a monolayer was formed.

Preparation of P. gingivalis LPS and Baicalin
Lyophilized LPS from P. gingivalis with type II fimA (strain code TDC60) was kindly provided by Prof. Y. Abiko (Nihon University, Japan). The LPS was prepared using the hot phenol water method [40,41]. It was reconstituted in Dulbecco's phosphate-buffered saline (DPBS) to a concentration of 1.0 mg/ml, followed by filtration through a 0.2 mm cellulose acetate membrane filter (Millipore, MA, USA). Baicalin powder (solvent extracted with a purity .95% as tested by HPLC) was obtained from the Hong Kong Jockey Club Institute of Chinese Medicine, Hong Kong. It was dissolved in pure dimethyl sulfoxide (DMSO) (Sigma), and then diluted in DPBS to 1.0 mM and finally filtered for sterilization. Working solutions were made with fresh oral keratinocyte medium on the experiment day.

Total RNA Extraction and cDNA Synthesis
Total RNA was extracted using the RNeasy mini kit (Qiagen, CA, USA). Briefly, cells were lysed with the buffer RLT, and the lysate was applied to an RNeasy Mini spin column. After several rounds of washes using the buffer RW1 and RPE, total RNA was bound to the column and other cell components were efficiently washed away. At the end, total RNA was eluted in RNase-free water. To avoid the contamination of genomic DNA, on-column DNase digestion was performed during RNA purification. The concentration of purified RNA was quantified by measuring its 260 nm UV absorbance on a NanoDrop spectrophotometer (Thermo, MA, USA). The integrity of purified RNA was evaluated by checking the ratio of 28 S rRNA and 18 S rRNA bands on an agarose gel. cDNA was synthesized using the Quantitect Reverse Transcription Kit (Qiagen). In brief, 1.0 mg of total RNA was pre-incubated with the gDNA Wipeout Buffer at 42uC for 2 min to remove any residual genomic DNA. The mixture was then incubated with the Quantiscript Reverse Transcriptase, Quantiscript RT Buffer and RT Primer Mix at 42uC for 30 min, followed by a termination step at 95uC for 5 min.

Real-time Polymerase Chain Reaction (PCR)
Each real-time PCR reaction mix contained 10.0 ml of the QuantiFast SYBR green master mix (Qiagen), 1.0 ml of cDNA template (5.0 ng), 1.0 ml of forward primer (10 mM), 1.0 ml of reverse primer (10 mM) and 7.0 ml of ultra-pure water. The reaction condition was set as follows: an initial activation at 95uC for 5 min, followed by 40 cycles at 95uC for 10 s and 60uC for 30 s. The primer sequences were: for IL-6, 59-AATCAT-CACTGGTCTTTTGGAG (forward), 59-GCATTTGTGGTTGGGTCA (reverse); and for IL-8, 59-GA-CATACTCCAAACCTTTCCACC (forward), 59-AACTTCTC-CACAACCCTCTGC (reverse); for b-actin (ACTB),59-TTGGCAATGAGCGGTT (forward), 59-AGTTGAAGG-TAGTTTCGTGGAT (reverse). All the primers were designed to amplify a region that lasts 100-250 base pairs long and contains at least one intron. They had passed our in-house amplification efficiency and specificity tests prior to usage. To check for nonspecific primer binding or co-amplification of residual genomic DNA, the melting curve was analyzed after each running. To detect foreign DNA contamination, a no-template control which contained all the reagents except the cDNA template was included in each running. Raw fluorescence data were analyzed by an Excel workbook called DART-PCR which automatically calculates threshold cycles, relative quantification values and amplification efficiencies [42].

Enzyme-linked Immunosorbent Assay (ELISA)
ELISA kits (R&D, MN, USA) were used to quantitatively determine the concentrations of IL-6 and IL-8 in culture supernatants. In brief, protein samples were pipetted into a microplate pre-coated with anti-IL-6 or anti-IL-8 antibodies and incubated at room temperature (RT) for 2 h. The plate was then washed three times with washing buffer to remove unbound samples. Subsequently, enzyme-linked polyclonal anti-IL-6 or anti-IL-8 antibodies were added and incubated at RT for 1 h. Following another three washes, a substrate solution was added and incubated at RT for 20 min. A blue color was then developed in direct proportion to the amount of the target cytokine in each well. Lastly, a stop solution was added to stop the color reaction. The absorbance was measured at 450 nm by a microplate reader (PerkinElmer, MA, USA).

Protein Extraction
Cytoplasmic and nuclear proteins were extracted using the NE-PER Nuclear and Cytoplasmic Extraction kit (Thermo). Two reagents, the Cytoplasmic Extraction Reagents I and II, were added to cell pellets to lyse cells. The cytoplasmic proteins released were collected by centrifugation. The remaining intact nuclei were lysed with the Nuclear Extraction Reagent and the nuclear proteins released were collected by centrifugation. The concentrations of fractionated proteins were measured by the BCA protein assay kit (Thermo).

Western Blot
Protein samples were separated on 10% SDS-polyacrylamide gels by electrophoresis and subsequently transferred to polyvinylidene difluoride membranes (Roche, IN, USA) by using the Mini-PROTEAN tetra electrophoresis system and Mini Trans-Blot transfer system (Bio-rad, CA, USA). Afterwards, the membranes were incubated with the Protein-Free T20 (TBS) Blocking Buffer (Thermo) at RT for 1 h and then probed with the primary antibodies (1:2000) at 4uC overnight with gentle agitation. On the next day, the membranes were washed and incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies at RT for 1 h. They were then washed again and incubated with the SuperSignal West Pico Chemiluminescent Substrate (Thermo) for 5 min. The signals of antigen-antibody complexes were developed on X-ray films. The density of the developed bands was quantified by the ImageJ software. The rabbit monoclonal antibodies (mAbs) against human IkBa, phospho-IkBa (serine32), phospho-p38 MAPK (Thr180/Tyr182), phospho-JNK (Thr183/Tyr185) and a-tubulin were obtained from Cellsignaling (MA, USA). HRPconjugated goat polyclonal antibodies against rabbit IgG were obtained from Thermo.

NF-kB p65 Transcription Factor Assay
A NF-kB p65 Transcription Factor Kit (Thermo) was used to measure the level of p65 transcription factor in nuclear protein samples. It contains a 96-well plate pre-coated with a biotinylated consensus DNA sequence which only binds p65. Briefly, nuclear protein samples were added to each well with binding buffer and incubated at RT for 1 h. The plate was then washed and incubated with primary anti-p65 antibody at RT for 1 h. Following another around of washing, secondary HRP-conjugated antibodies were added to the plate and incubated at RT for 1 h. Lastly, a chemiluminescent substrate solution was added to the wells. The signal image was captured with a CCD camera and the signal intensity was measured by a multiplate reader.

PCR Array
A panel of 89 genes associated with TLR signal transduction was investigated simultaneously using the RT 2 Profiler TM PCR Arrays (SAbiosciences, MD, USA). RNA samples were firstly reverse transcribed into cDNA templates by the RT 2 First Strand Kit (SAbiosciences). The diluted cDNA templates were subsequently mixed with the RT 2 qPCR Master Mix (SAbiosciences) and H 2 O. 25 ml of the mixture were loaded into each well of the array plate which contained pre-coated specific primers. The realtime PCR was performed as follows: an initial incubation at 95uC for 10 min, then 40 cycles at 95uC for 15 s and 60uC for 1 min. Data analysis was undertaken by using the SAbiosciences webbased PCR array data analysis software.

Statistical Analysis
All experiments were repeated three times. The data were presented as mean6SD and the statistical significance was evaluated by one way ANOVA using the SPSS 16.0 software. A p-value,0.05 was considered statistically significant.

Baicalin Downregulated P. gingivalis LPS-upregulated Expression of IL-6 and IL-8
HOKs were pre-treated with baicalin (0-80 mM) for 24 h, and subsequently treated with fresh media containing P. gingivalis LPS (10 mg/ml) with or without baicalin (0-80 mM) for 3 h. The culture supernatants and total RNA were collected for ELISA and real-time PCR analyses, respectively. We discovered that baicalin at 40 mM and 80 mM significantly suppressed P. gingivalis LPSupregulated IL-6 mRNA expression (Fig. 1A); and baicalin at 10 mM, 20 mM, 40 mM and 80 mM significantly downregulated P. gingivalis LPS-upregulated IL-6 protein expression (Fig. 1B). Baicalin at 80 mM also significantly suppressed P. gingivalis LPSupregulated IL-8 mRNA and protein expression ( Figs. 2A & 2B). As baicalin at 80 mM contained 0.08% DMSO, the observed downregulation could have been partially caused by DMSO. To exclude this possibility, a vehicle control group was set up by treating cells firstly with 0.08% DMSO for 24 h, and then with P. gingivalis LPS (10 mg/ml) and 0.08% DMSO for 3 h. No DMSOmediated inhibition on IL-6 or IL-8 expression was found (Figs. 1 and 2).

Baicalin Displayed Inhibitory Effect on P. gingivalis LPSinduced Activation of NF-kB, p38 MAPK and JNK
In resting cells, inactive NF-kB (p65/p50) is retained in the cytoplasm by an inhibitory protein called IkBa [43]. Upon stimulation, IkBa is ubiquitinated and degraded by 26 S proteasome, resulting in the translocation of NF-kB to the nucleus where it binds to the target genes and initiates gene transcription [43]. As NF-kB plays a central role in P. gingivalis LPS-mediated cell response and the expression of IL-6 and IL-8 is dependent on NF-kB signaling [44], we were interested to exam whether baicalin could have any inhibitory effects on P. gingivalis LPS- , and sequentially stripped and re-probed with rabbit anti-phospho-p38 MAPK mAbs (1:2000), rabbit anti-phospho-JNK mAbs (1:2000), and rabbit anti-IkBa mAbs (1:2000). For loading control, the membrane was probed with rabbit anti-a-tubulin mAbs (1:4000). B. The densitometry analysis of the signals. Cells treated with culture media alone served as the blank control group, and those treated with P. gingivalis (P.g.) LPS (10 mg/ml) alone represented the positive control group. Cells treated with P.g. LPS at 10 mg/ml and 0.08% DMSO served as the vehicle control group. Data of three independent experiments were depicted as relative fold change as compared with the blank control group (set as 1). For the p-JNK protein, the positive control group (LPS) was set as 1 since the signals of the blank control group at 15 min and 30 min were undetectable. *p,0.05 and **p,0.01 as compared with the respective positive control group (LPS) at each time point. BI: baicain. doi:10.1371/journal.pone.0051008.g003 activated NF-kB. Cells were pre-treated with baicalin (80 mM) for 24 h, and thereafter treated with fresh media containing P. gingivalis LPS (10 mg/ml) with or without baicalin (80 mM) for 15, 30, and 60 min. A vehicle control group was set up by treating cells firstly with 0.08% DMSO for 24 h, and then with P. gingivalis LPS (10 mg/ml) and 0.08% DMSO for 15, 30, and 60 min. As shown in Fig. 3, baicalin significantly inhibited to different extents P. gingivalis LPS-induced phosphorylation of IkBa, p38 MAPK and JNK which act as the downstream of TLR2/4 signaling pathways [45].

Baicalin Suppressed P. gingivalis LPS-induced Nuclear Translocation of p65
The effect of baicalin on P. gingivalis LPS-induced nuclear translocation of p65 was examined by using a p65 transcription factor kit. Cells were pre-treated with baicalin (80 mM) for 24 h, and then treated with fresh media containing P. gingivalis LPS (10 mg/ml) with or without baicalin (80 mM) for 15, 30, and 60 min. Compared with P. gingivalis LPS-treated samples, baicalin succeeded to suppress the amount of translocated p65 in the nuclear protein extracts at 60 min (Fig. 4).

Baicalin Modulated P. gingivalis LPS-induced Expression of Genes Associated with TLR Signaling
Lastly, a PCR array assay was undertaken to profile the expression of genes associated with TLR signaling. Cells were pretreated with baicalin (80 mM) or culture media for 24 h, and then treated with fresh media containing P. gingivalis LPS (10 mg/ml) with or without baicalin (80 mM) for 3 h. The total RNA was purified and reverse transcribed into cDNA templates. The templates used in PCR array were pooled equally from three independent experiments. Compared with the P. gingivalis LPStreated cells, the expression of CCL2, CSF2, CSF3, CXCL10, IL-8, V-fos FBJ murine osteosarcoma viral oncogene homolog (FOS) and interferon, beta 1, fibroblast (IFNB1) was significantly downregulated over two folds in baicalin/P. gingivalis LPS treated . Baicalin suppresses P. gingivalis LPS-induced nuclear translocation of p65. A. The representative experiment was performed by pooling nuclear protein extracts equally from three independent experiments. 2 mg aliquots were added to each well. The assay was carried out according to the manufacturer's instruction. B. The intensity analysis of the luminescent signals. Cells treated with culture media alone served as the blank control group, and those treated with P. gingivalis (P.g.) LPS (10 mg/ml) alone represented the positive control group. Cells treated with P.g. LPS at 10 mg/ml and 0.08% DMSO served as the vehicle control group. Data from three independent experiments were depicted as relative fold change as compared with the blank control groups (set as 1). *p,0.05 and **p,0.01 as compared with the respective positive control group (LPS) at each time point. BI: baicalin. doi:10.1371/journal.pone.0051008.g004 cells (Table 1). Notably, CCL2, CSF3 and CXCL10 were markedly downregulated over four folds. On the other hand, other genes including cluster of differentiation 80 (CD80), interleukin-1 receptor-associated kinase 1 (IRAK1), jun protooncogene (JUN), TLR6, Ubiquitin-conjugating enzyme E2 variant 1 (UBE2V1), sterile a and TIR motif-containing 1 (SARM1), single Ig IL-1-related receptor (SIGIRR), TIR domain containing adaptor protein (TIRAP) and TLR1 were significantly upregulated over two folds in baicalin/LPS-treated cells (Table 1).

Discussion
Periodontal disease results essentially from the consequence of a disrupted immuno-inflammatory homeostasis of bacteria-host interactions [5]. In susceptible individuals, when host response fails to limit and resolve early infection timely, cytokine expression may become dysregulated and destructive to tissues [46,47]. As IL-6 is a stimulator of bone resorption and IL-8 is a potent neutrophil chemoattractant and activator [48,49], prolonged and excessive production of these pro-inflammatory cytokines could contribute to periodontal tissue damage. Our present study shows that baicalin could significantly downregulate P. gingivalis LPS-upregulated production of IL-6 and IL-8 in HOKs. This observation goes in line with the concept of host modulatory therapy, suggesting that baicalin could potentially be used for modulation of host response in treatment of periodontal disease.
The present study also reveals that baicalin may inhibit P. gingivalis LPS-induced activation of NF-kB, p38 MAPK and JNK. Due to their involvements in a variety of human diseases, NF-kB, p38 MAPK and JNK have become therapeutic targets and several NF-kB inhibitors have been discovered, such as sulindac [50], IKK inhibitor [51,52] and resveratrol [53]. It has been shown that SD828, a p38 MAPK antagonist, could suppress LPS-induced alveolar bone loss in periodontitis rats [54], and JNK inhibitors like CEP-1347 and AS601245 exhibit protective effects on neurons [55,56]. In the present study, the exact mechanism of baicalin-induced inhibition of P. gingivalis LPS-upregulated expression of IL-6 and IL-8 in HOKs remains to be further elucidated. While it could be speculated that the inhibition observed could have been exerted directly on IKK, p38 MAPK and JNK; or on the upstream kinases such as transforming-growth-factor-bactivated kinase 1 (TAK1) (kinase of IKK and p38/JNK MAPK) [57][58][59], interleukin-1 receptor-associated kinase 1 (IRAK1) (kinase of TAK1), or IRAK4 (kinase of IRAK1) [60][61][62].
According to the PCR array results, CCL2 [63], CSF3 [64,65] and CXCL10 [66] were greatly downregulated over four folds by baicalin treatment. The transcription of CCL2 and CSF3 is regulated by NF-kB [67,68]. In response to LPS, CXCL10 is induced in a TLR4-mediated MyD88-independent pathway [15,69,70]. The exact reasons that baicalin could downregulate both LPS-induced MyD88-dependent and MyD88-independent genes remain unclear. The possible mechanisms are as follows: i) baicalin might enable to interfere with the binding of P. gingivalis LPS to TLR4; ii) it could inhibit multiple downstream kinases of TLR4 signaling, such as IKK, TAK1 and TANK-binding kinase 1 (TBK1) (kinase of IRF3) [71]; iii) as the optimal transcription of CXCL10 requires a coordinated binding of activated IRF3 and NF-kB to the promoter region [15,69], baicalin-mediated inhibition of NF-kB could have interfered with the expression of CXCL10. Further study is warranted to clarify these points.
Over the last three decades, the growing knowledge of periodontal pathogenesis has appreciated the crucial role of host response in the initiation and development of periodontal disease. Recently, TLR signaling has become an attractive target of host modulation therapy due to its central role in activating immunoimflammatory response in the development of periodontitis [72]. To date, a number of negative regulatory strategies for overactivated TLR signaling have been proposed, such as natural/ synthetic antagonists [73,74], BB-loop peptides [75], miRNA [76] and kinase inhibitors [77]. Here, we report for the first time that baicalin can significantly downregulate P. gingivalis LPS-upregulated IL-6 and IL-8 expression in HOKs, through negative regulation of TLR signaling. Based on these findings, baicalin may potentially serve as a host response modulator in the control of periodontal disease by negative regulation of TLR signaling. Further clinical study is warranted to investigate the effectiveness of baicalin as a potential adjunct in treatment of patients with inflammatory diseases like periodontal disease.