The Anti-Inflammatory Effect of Zhibaidihuang Decoction on Recurrent Oral Ulcer with Sirt1 as the Key Regulatory Target

The syndrome of ROU is generally manifested as obvious pain, redness, and swelling of local ulceration area, accompanied by flushed face, red eyes, sore throat, and swollen gums. Traditional Chinese medicine (TCM) doctors believe that “yin deficiency” is one causative factor of ROU. Zhibaidihuang decoction (ZBDHD) is a prescriptively developed receipt, where Anemarrhena asphodeloides and Phellodendri amurensis Cortex are added in the original Liuweidihuang decoction. It is generally used for “yin deficiency” treatment. It can effectively reduce the recurrence of oral ulcers and release the severity of the disease. However, the mechanism of this activity remains to be elucidated. In this study, we found that ZBDHD has a certain therapeutic effect on the pathological changes of oral mucosa. Furthermore, the results of serum metabolomics showed ZBDHD influenced the synthesis and metabolism of certain fatty acids. The results of western blot, immunochemical, and immunofluorescence staining indicate that ZBDHD could increase the expression of Sirt1 and Foxp3 and suppress the expression and acetylation of NF-κB in oral mucosa cells. By screening active ingredients in ZBDHD, we found berberine, as well as other compounds, presenting high fitness of the Sirt1 reactive centre. Therefore, it is possible that ZBDHD can regulate the Sirt1-NF-κB pathway to improve fatty acids metabolism in the body, thereby achieving the effect of treating ROU.


Introduction
Recurrent oral ulcer (ROU) is one of the most common oral mucosal diseases and its incidence rate is 30% [1]. Come so far, the mechanism is still unclear; most studies indicate that ROU is associated with immune inflammation [2]. It was considered that the ROU included five stages including initiation, primary damage response, expanded damage response, ulcer formation, and repair, while inflammatory response was throughout the process. e clinical research showed that CD4 + , CD25 + , and Foxp3 + were significantly reduced [3]. On the contrary, IL-6 and TNF-α concentration was significantly increased in peripheral blood of recurrent ulcers patients [4]. Animal experiments showed that the expression of NF-κB p65 [5] and TNF-α significantly increased in recurrent oral ulcer model rats [6]. ese previous reports demonstrated that ROU is related to immune response and inflammatory response.
e Chinese medicine theory mentions that the unevenness of "yin" (cold) and "yang" (hot, fire) in the human body could result in diseases. TCM doctors considered that excess "yang" (fire) is an important factor in causing ROU in clinic [7]. e excess fire could be due to "yinxu," which means "yin deficiency". Fuzi (Radix Praeparata Aconitum carmichaelii Debx), Ganjiang (Rhizoma Zingiber officinale), and Rougui (Cortex Cinnamomum cassia) decoction (AZC) is one of the prescriptions for hyperthermia, used to treat "cold" syndrome. However, overdose of AZC could break the balance between "yin" and "yang," which produces heat and inflammation [8]. Previous work indicated that AZC could induce excess of "fire" to cause oral ulcer [9,10].
Zhibaidihuang decoction (ZBDHD), a traditional Chinese herbal receipt, is usually used to cure ROU and hyperactivity of "fire" due by "yin" deficiency in clinic [11,12]. e receipt was composed of eight herbs including Zhimu (Rhizoma Anemarrhena asphodeloides), Huangbai (Phellodendri amurensis Cortex), Shudihuang (Radix Rehmanniae Preparata), Shanzhuyu (Fructus Cornus officinalis), Shanyao (Radix Dioscorea opposita), Fuling (Sclerotium of Poria cocos), Mudanpi (Cortex Paeonia suffruticosa), and Zexie (Rhizoma Alisma orientale). In this prescription, Fuling, Mudanpi, and Zexie have capability of eliminating excess "yang" (fire). And Huangbai and Zhimu show activity on enhancing the efficiency. e active components in this receipt are very complex. is is due to the complex mechanisms of ZBDHD treatment. It was identified with modern biological method that AZC can influence the AMPK/Sirt1 pathway [10], as well as expression of NF-κB, one of inflammatory response cytokines [13]. Previous studies proved that Sirt1 was correlated to cell metabolism [14,15]. And Sirt1 is the key cytokine, which regulates the transcription of NF-κB [16,17]. erefore, in this research, we evaluated the regulation mechanism of ZBDHD treatment on oral mucosa through Sirt1/NF-κB pathway with a AZC treated ROU rat model.

Animals.
Fifty healthy (180∼200 g) female SD rats were purchased from Shanghai SLAC Lab. Animal Ltd. (Shanghai, China) and housed in laboratory animal research centre of Zhejiang Chinese Medical university. All animal experiments have passed the ethics committee with accepted Nr. ZSLL-2016-116.

Decoction Preparation.
All of the herbs were purchased from Zhejiang Province Famous Doctor Pavilion and Zhejiang Chinese Medical University Medical Slicing Factory. e preparation of AZC decoction refers to our previous publication [8]. ZBDHD preparation was also described previously [13] and was diluted to specific concentration for experiments. And the decoction concentration in this work presented as milli gram herbal mixture per milli liter.

Animal Treatment.
In this experiment, six rats (control group, Ct), were given distilled water. For model preparation, the oral mucosa from twenty healthy SD rats was mixed with complete Freund's adjuvant (1 : 1 v/v) and sterile-filtrated with a sterile filter (0.22 μm diameter). Twenty-four rats were subcutaneously injected 0.1 mL this mucosa emulsion on both sides of back once a week. And rats were internally orally injected 0.1 mL mucosa emulsion, also. Simultaneously, rats were given 4 mL 1 g/mL AZC by intragastric administration (i.g.) every day for 4 weeks. For ZBDHD protective efficiency evaluation, these twenty-four rats were separated into four groups. One group (model group, AZC) was given 4 mL/day distilled water by intragastric administration (i.g.). Simultaneously, the other three groups were treated with 4 mL/day high dose (2 g/mL), middle dose (1 g/mL), and low dose (0.5 g/mL) ZBDHD by i.g. for a week, respectively, which were marked as ZBDHD-H, ZBDHD-M, and ZBDHD-L.

Histological Study.
e 10% formalin solution fixed oral mucosa (of each rat in different groups) was embedded in paraffin which was cut and stained with Hematoxylin & Eosin (H&E) stain for electron microscope (Carl Zeiss, Germany) observation.

Western
Blot. e expressions of signalling molecules, Sirt1, NF-κB p65, Stat3, and Foxp3 in oral mucosa cell from different group rats were comparatively identified with western blot. e oral mucosa was sheared in lysate buffer by tissue homogenizer, and cell debris was discharged by centrifugation. e protein concentration in cell lysate was quantified with Bradford. 20-30 μg proteins were mixed with loading buffer and separated with an SDS-PAGE gel. For western blot analysis, the proteins in gel were electrically transferred on a nitrocellulose filter membrane at 80 V for 2 hours in a transfer buffer system (3.03 g Tris + 14.4 g glycine + 200 mL methanol in ddH 2 O to a final volume of 1000 mL). Before hybridization, the membrane was cut according to the mass of detected proteins and then incubated with antibodies specific for β-actin, Sirt1, NF-κB p65, Stat3, or Foxp3 separately at 4°C overnight. On the next day, membranes were washed and incubated with the respect secondary antibody for 2 hours. e blots were exposed and scanned using an Odyssey Infrared Imaging system (LI-COR Biosciences), and the blots were quantified using the Odyssey Infrared Scanning system software.

Statistical Analysis.
All the data in this study were expressed as mean ± standard deviation (SD). And t-test was used to determine the differences between control and treatment groups with SPSS 15.0. e statistically significant difference was considered at level P < 0.05.

Pathological Changes of Oral Mucosa Induced by AZC and
Proactive Effect of ZBDHD. In 4 weeks, the oral mucosa of the model group exhibited irritation and inflammation, which is shown in Figure 1(a) and indicated by a black arrow. e aberrant hyperplastic epithelium and infiltrating of inflammatory cells could be observed in H&E strained oral mucosa of AZC model rats (Figure 1(c)), but not in control group rats (Figure 1(b)). In comparison with that of rats in AZC model group, oral mucosa of ZBDHD (high dose and middle dose) (Figures 1(d) and 1(e)) treated rats showed clear cell structure and improved of inflammation, obviously. However, no positive effects could be observed in ZBDHD (low dose) group (Figure 1(f )). e changes of body weights during the last 7 days treatment were monitored (Table 1). Although those of the ZBDHD treated rats were larger than those of model rats, no significant differences were observed.

ZBDHD Influences the Metabolites Variation.
As a part of system biology, metabolomics is a new discipline which can demonstrate the curative effect of TCM therapy. In this study, metabolic profiles of serum samples from rats in control and ZBDHD-H (Figure 2(a)) groups were analyzed by HPLC-QTOF/MS. e data were discriminated with the principal component analysis (PCA) (Figure 2(b)), partial least squares discriminant analysis (PLS-DA) (Figure 2(c)), and the orthogonal partial least squares-discriminate analysis (OPLS-DA) (Figure 2(d)). e metabolites were considered as potential biomarkers responsible for the metabolic profile, when P value of t-test <0.05 and VIP >1.

e Effect of ZBDHD on Sirt1/NF-κB Pathway in Oral
Mucosa. Sirt1 is a protein deacetylase that interferes with the NF-κB signaling pathway, which regulates cell anti-inflammatory function. erefore, the expression of cytokines, Sirt1, NF-κB, and TNF-α, in oral mucosa of rats in different groups, was immunohistochemically imaged in brown ( Figure 3). In the cell of model rats, the expressed Sirt1 was obviously less than those of control rats. On the contrary, the expressed NF-κB and TNF-α were more than those of control group. e expression of Sirt1, NF-κB, and TNF-α in oral mucosa cells of rat was ameliorated by ZBDHD treatment. Simultaneously, the location of Sirt1 (red stained) and deacetylated NF-κB (green stained) in cells was studied with immunofluorescence detection (Figure 4). e yellow signal could be obtained in merged photos. In comparison, in the model group (Figure 4 Merge) more yellow signal could be observed than in the other two groups. To confirm this result, the expression of Sirt1 and NF-κB in cells was studied by western blot, as well as Foxp3 and Stat3. e high dose ZBDHD treatment (2 g/mL) could inhibit the expression of NF-κB and stimulate the expression of Sirt1, significantly ( Figure 5). In addition, the expression of Foxp3 was also influenced by the treatment in a dose-dependent manner. However, the expression of Stat3 was stimulated by ZBDHD but not significantly.
Evidence-Based Complementary and Alternative Medicine e position of cell nucleus, acetyl lysine, and NF-κB was, respectively, hybridized with blue, red, and green fluorescein labeled specific antibodies. e yellow spot, in merged image, indicated the co-location of acetyl lysine and NF-κB (red arrow marked).  (Table S2). e interaction specification was evaluated by affinity score. And eight of the most sensitive candidates are shown in Table 3, according to their affinity score.

Discussion and Conclusion
Herbal medicines have been used to treat various diseases for many centuries, in Chinese society. Zhibaidihuang  Note: the data are expressed as mean ± SD (n � 6). 4 Evidence-Based Complementary and Alternative Medicine decoction (ZBDHD), a poly herbal formula, has been used to treat syndrome of hyperactivity of "yang" (fire) due to "yin" deficiency for thousands of years [21]. Recurrent oral ulcers are a typical symptom of hyperactivity of "fire" due to "yin" deficiency [22]. In modern medicine, oral ulcer is a kind of inflammatory diseases, which occurs due to a damage in epithelium induced by many certain factors, such as hormonal changes [23], lack of essential vitamins [24], and     disorder of metabolism which could induce oral ulcer [25]. It is validated by clinical trial that ZBDHD treatment could promote the ulcer healing [26] and improve the fitness of recurrent oral ulcer significantly [13]. Our previous research shows the extracts of AZC herbs may induce "fire" syndrome by influencing Treg cells and immunosuppressive cytokines [8]. In this research, AZC combined with immunological method induced immune dysfunction and oral ulcer which was similar to ROU clinical symptoms. Histopathological studies showed the cell inflammation infiltration and mucosal surface damage in model rats (Figures 1(a) and 1(c)). e protection capability of ZBDHD was studied in this research. Mucosal surface damage in model and low dose ZBDHD treated rats could be observed under microscope by H&E staining, but the rats treated with high and middle dose ZBDHD showed a sign of healing which was indicated the absence of damage ( Figure 1). Simultaneously, the difference of metabolites between model group and high dose ZBDHD treatment group was evaluated by metabonomic study which indicated that ZBDHD treatment influenced synthesis and metabolite of some fatty acids, such as phytosphingosine and palmitic amide ( Table 2). High dose ZBDHD treatment attenuated AZC induced distractive expression of cytokines, Sirt1, NF-κB, and TNF-α (Figures 3 and 4). e results of our previous study also indicated that AZC can stimulate the expression of NF-κB [13]. e expression of cytokines in Sirt1 signaling pathway was quantitatively analyzed by western blot ( Figure 5). Sirt1, which belongs to Sirtuins family, is involved in many human physiological processes, such as aging, DNA repair, cell apoptosis [27], and inflammatory response [28]. Sirt1 catalyzes the deacetylation of Lys 310 s in p65 sigmasubunit of NF-κB [29], for suppression the regulation  Evidence-Based Complementary and Alternative Medicine activity of NF-κB [30]. To predict the phytochemistry and pharmaceutical mechanism of ZBDHD, we used the bioinformatic method to screen TCMSP database (http:// lsp.nwu.edu.cn/tcmsp.php). A total of 78 ingredients present Sirt1 interaction potential. Eight of the 78 show relative high binding affinity with reactive centre of Sirt1 (Table 3). ey are coptisine, berberine, sitosterol, diosgenin, mairin, anemarsaponin F_qt, trametenolic acid, and alisol B monoacetate.
ey have various pharmaceutical activities. For example, alisol B monoacetate has anti-tumor, anti-allergic, and inhibiting infection from hepatitis B virus [31]. Trametenolic acid can improve cerebral ischemia [32]. Anemarsaponin inhibits inflammation by inhibiting the phosphorylation of NF-κB p38 [33]. Mairin, also called betulinic acid, has strong effects on anti-HIV, anti-inflammatory, anti-diabetic, and anti-microbial activities [34]. Diosgenin has anti-cancer, anti-diabetic, anti-coagulation, anti-thrombosis, antiinflammatory, anti-viral, and anti-aging properties [35]. Sitosterol has a strong effect on decreasing blood fat and cholesterol [36]. Coptisine and berberine present higher score than others. Both ingredients are the structurally related isoquinoline alkaloids. But the results of previous studies indicated their differential functions. Coptisine, an antioxidant [37], could induce autophagy to exert anti-cancer effects [38]. No references indicated the anti-inflammatory activity of coptisine. In comparison, berberine was reported to have a strong effect on antiinflammatory, previously [39]. As a powerful supplement with many benefits, berberine, one of the eight candidates, showed high binding affinity with the amino acid residues in active centre of Sirt1 ( Figure 6). To our knowledge, it is the first time to evaluate the binding site of berberine on Sirt1.
In the present study, we used high dose "hot" herbs potions combined with immunological method to induce ulcer in mouth of model rats. In 14 days, the visible cavity in mucous membrane and decrease of body weight were observed in model rats, in comparison with the other groups. e phenomenon was identical to that in clinic [40]. Serum metabolomics analysis results showed that the metabolisms of some lipids, such as phytosphingosine, palmitic amide, and MG (0 : 0/16 : 0/0 : 0), were significantly inhibited by ZBDHD. Phytosphingosine [41], a phospholipid, is a major component of all biological membranes and sphingolipid metabolites. e palmitic amide is one of primary fatty acid amides, which are a kind of signaling molecules [42]. However, the function of palmitic amide is less known. MG (0 : 0/16 : 0/0 : 0) is a monoacylglyceride, consisting of a fatty acid chain covalently bonded to a glycerol molecule through an ester linkage. It is an intermediate in glycerophospholipid metabolism process [43]. On the contrary, ZBDHD treatment upregulated the serum concentrations of α-linolenic acid and LysoPC (20 : 4 (5Z, 8Z, 11Z, 14Z)), LysoPC (20 : 3 (5Z, 8Z, 11Z), and LysoPE (0 : 0/20 : 0). Both LysoPC (20 : 4 (5Z, 8Z, 11Z, 14Z)) and LysoPC (20 : 3 (5Z, 8Z, 11Z)) are lysophosphatidylchoine. LysoPE (0 : 0/20 : 0) are lysophosphatidylethanolamine. ey are structural compounds in cell membrane at a low density. us, they are biomarkers of cell proliferation [44]. In addition, LysoPC and LysoPE are both post-injury marker metabolites [45]. LysoPC was a strong chemoattractant of T-lymphocytes, which can promote antibody formation and macrophages stimulus [46]. LysoPE is mobilizer of intracellular Ca 2+ in some cell types [47,48]. Calcium is one of the initial triggers in immune response for wound healing [49]. α-Linolenic acid served as substrate for unsaturated fatty acid synthesis, which showed anti-inflammatory effects [50]. erefore, the results of metabolomics analysis indicated that the ZBDHD can reduce inflammation and promote the ulcer healing.
It had been reported that Sirt1, the silent information adjustment factor 2 related enzyme I, could suppress fatty acid metabolism [51]. In addition, Sirt1 can catalyze acetylation of transcription factor Forkhead Box P3 (Foxp3), an inhibitor of NF-κB [52]. NF-κB is an important pro-inflammatory factor and could influence the expression of inflammatory factors, such as TNF-α, IL-1β, and IL-8 [53]. erefore, we hypothesized that Sirt1 could be one of the target factors of ZBDHD treatment. Evidence-Based Complementary and Alternative Medicine

Conclusion
In present study, we demonstrated that ZBDHD treatment upregulated the expression of Sirt1 and attenuated the overexpression of TNF-α. ese results confirmed that Sirt1 may be the main target of ZBDHD to reduce the mucosa inflammation. ZBDHD could improve the symptoms of oral ulcer in model rats. And the capability mechanism may be related to Sirt1-NF-κB pathway, immune regulation, and metabolism regulation. And the berberine, one of the constituents in ZBDHD, is a key regulator of Sirt1 (Figure 7).

Data Availability
All data used to support the findings of this study are included within the article, and these data can also be made accessible on website https://figshare.com/s/3ba0eaf51b3c2c78bc7b.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
All the authors provided critical feedback and contributed to research designation, data collection, and analysis, respectively. In detail, Ms Yajie Shao and Shanshan Lei performed the experiments and collected data. Associate professor Bin Ding analyzed the data and drafted this manuscript. Ms Jinjun Ji was responsible for lab management and helped with laboratory procedure. Dr Yu Dong was responsible for serum metabolomics analysis. Professor YS Fan designed the animal model preparation and financially supported this study. Mr Weihong Ge helped in modifying and reediting this manuscript. Professor L Xu contributed to the conception of the whole work and financially supported this work.