Protective Effects of Clinacanthus nutans (Burm.f.) Lindau Aqueous Extract on HBV Mouse Model by Modulating Gut Microbiota and Liver Metabolomics

Background Clinacanthus nutans (Burm.f.) Lindau (C. nutans) has been used in the therapy of hepatitis B (HB) and is effective; however, the mechanism of action has not been elucidated. Objective To investigate the protective effects of C. nutans aqueous extract on the hepatitis B virus (HBV) mouse model based on correlation analysis between gut microbiota and liver metabolomics. Materials and Methods We firstly constructed the animal model by high-pressure injection of pcDNA3.1(+)/HBV plasmid into the tail vein and treated it with C. nutans. The biomarkers and inflammatory cytokines of HB were detected by enzyme-linked immunosorbent assay and quantitative PCR; the Illumina-MiSeq platform was used for investigating gut microbiota; the LC-MS/MS method was utilized on screening liver tissue metabolites; multiomics joint analysis was performed using the R program. Results Compared with the modeling group, C. nutans significantly decreased the expression levels of HBsAg, IL-1β, TNF-α(P < 0.05) in the serum, and cccDNA (P < 0.05) in the liver tissues of mice. C. nutans dramatically reduced the ratio of Firmicutes and Bacteroidetes (P < 0.05) and significantly declined the proportion of Lactobacillaceae and Lactobacillus(P < 0.05), dramatically increasing the relative abundance of Bacteroidales_S24-7_group, Rikenellaceae, and Alistipes(P < 0.05); LC-MS/MS analysis results showed that C. nutans dramatically upregulate hippuric acid, L-histidine, trehalose, D-threitol, and stachyose and downregulate uridine 5′-diphosphate, cholic acid, trimethylamine N-oxide, CDP-ethanolamine, and phosphorylcholine (P < 0.05). The correlation analysis revealed that C. nutans affects the related metabolite levels of hippuric acid and cholic acid through the modulation of crucial bacteria (Alistipes) (P < 0.01), exerting specific anti-inflammatory effects. Conclusion These results suggest that C. nutans exerts protective effects in HBV model mice, showing the therapeutic potential for anti-HBV infection.


Introduction
HBV infection, which remains a severe public health problem [1], causes over 400 million people to be infected and 1 million patients to die each year [2]. Interferon and nucleoside analogs are used for treatment. However, these approved therapeutic agents are still not reliable [3]. Terefore, new strategies are needed to make the therapeutic efect better. Presently, the induction of the HBV mice model by using a hydrodynamically injected method to simulate the natural process of human infection could be a mature technology for new drug research of HBV [4].
C. nutans is a plant that belongs to the family of Acanthaceae, possessing potential and diverse medicinal values in tradition for snake and insect bites, skin rashes, herpes simplex virus, and varicella-zoster virus lesions [5]. Te extract of C. nutans has various biological efects, such as anti-infammatory, antioxidant, immune response, antimicrobial, and antidengue activity [6].
Gut microbiota has excellent signifcance for some physiological and pathological processes. Healthy fora has characteristics of richness, diversity, and stability of the intestinal ecosystem and can self-regulate and recover after being disturbed. Dysbiosis refers to the changes in gut microbiota driven by a series of factors, including the composition and function aspects, which exceed the resistance and resilience of microorganisms themselves [7]. Gut microbiota disorders are highly correlated with HB [8].
Research shows that gut microbiota changed signifcantly between HB patients and healthy ones [9]. Te use of fecal microbiome transplantation can promote the HBeAg clearance rate of patients after persistent antiviral therapy [10]. In mouse models, an imbalance of gut microbiota may damage HBV-specifcT-cell response and prolong the course of the disease [11].
Te interaction between gut microbiota and liver diseases has aroused widespread interest. Bile acids, one of the hepatic metabolites, afect the intestinal secretion system of microorganisms through the hilum and bile [12][13][14]. Sterile mice can prevent the pathogenic role of intestinal microorganisms in viral hepatitis [15]. In patients with HBV infection, the imbalance of the microbiome promotes the increase of cytokine secretion, which induces the malignant progress of chronic infammation and liver lesions [16,17].
Persistent hepatitis can seriously damage gut microbiota and liver metabolism function [18]. At the same time, abnormal metabolites are also highly correlated with liver diseases. High levels of aromatic amino acids in serum may be a signifcant marker of liver tissue dysfunction and the pathogenesis of chronic liver diseases [19][20][21][22][23]. Studies have found that gut microbiota regulates the host's metabolic phenotype [24,25]. Terefore, elucidating the regulation of gut microbiota and metabolite characteristics of HB will help reveal the further mechanism of the efcacy of C. nutans and lay a foundation for the development of potential medicinal plants.

Induction of the Animal Model and Drug Administration.
Te plasmid pcDNA3.1(+)/HBV (provided by Dr. Xin Wang, Ocean University of China, Qingdao, China) was used in this study. 10 μg of plasmid DNA was diluted with physiological saline according to 0.1 mL/g of the mouse body weight, and then the solution was injected into the tail vein of the model group mice within 10 s. Six-week-old male mice were hydrodynamically injected (HI) with plasmids [26]. Te control group mice were HI with physiological saline as the same protocol. After 24 hours of HI, the serum of mice was collected; then, we used ELISA kits (Kehua Bioengineering Co. Ltd., Shanghai, China) to detect ALT and HBsAg levels, each sample was preprocessed by diluted 1 : 10 with PBS. Te modeling mice were selected and distributed into the C. nutans aqueous extract group (YDC), the modeling group (MOD), and the entecavir solution group (ETC), with 10 mice per group. And 10 mice of the control group were selected as the nondiseased group (NOR). All animals weigh approximately. Te group of YDC and ETC mice were orally administered 0.2 mL reagent (as described in the preparation of plant extract) or 0.2 mL entecavir. We calculated the equivalent doses for humans and mice, a group of YDC mice was gavaged 1.95 g/kg/d C. nutans for 10 days, a group of ETC mice was gavaged 3.2 mg/kg/ d entecavir, and the other group of mice was gavaged with the same volume of distilled water.

Measurement of Biomarkers and Infammatory
Cytokines of HB. On day 10, all animals were sacrifced, then the blood samples were collected in sterile Eppendorf tubes by the method of blood taking from the retroorbital vascular plexus of mice, and centrifugated at 2000 g for 10 mins to separate serum. After that, the liver tissue was excised on the ice and cut into two equal parts, fast stored in liquid nitrogen until use. Te HBsAg, IL-1β, and TNF-α in the serum were detected in the same protocol as the preceding description. Versus ∼1 gm of the liver tissue was taken and cut into pieces, we added 300 μL DNA Dentaured&Fragmented bufer with 10% protease K ( Solarbio Bio-Engineering Co., Ltd., Shanghai, China.) in each tube to resuspend the tissue homogenate. Ten, incubated it at 55°C for 2 h and centrifugated for 5 mins (at 12,000 g and room temperature). After that, we added 300 μL DNA extract reagent. We added 300 μL isopropanol, mixed it upside down, set it at −20°C for 1 h, and centrifugated it for 5 mins again. Discarded the supernatant and added 300 μL 75% ethanol, mixed it well, and centrifugated it for 5 mins. Added 50 μL ddH 2 O after discarding the supernatant and stored at 4°C overnight. Te next day, Plasmid-Safe ™ ATP-dependent DNase was used to digest DNA. Finally, the relative contents of cccDNA were determined with mouse GAPDH expression levels. Primers are as follows:  [27]. Pretreatment of specimens, PCR experiments, and product purifcation were carried out by the operation manual. After that, we used the Miseq platform to perform deep sequencing. Te raw data were trimmed using Illumina analysis pipeline version 2.6. Te dataset was then analyzed using QIIME. Te clustering of operational taxonomic units (OTUs) for calculating richness and diversity indices is as reported in [28]. All sequences were classifed by the ribosomal database project (RDP) classifer tool [29]. Te protocol of examining the similarity between diferent samples, evaluating the distances between microbial communities from each sample, and comparing the membership and structure of communities in diferent samples was followed as described previously [30,31].

Metabolomics of Liver Tissue Analysis Using the UPLC-Q-TOF/MS Method and Correlation Analysis.
Te experimental site is in Shanghai Applied Protein Technology Co., Ltd, Shanghai, China. Te liver tissue samples were collected in sterile Eppendorf tubes and kept at −80°C until analysis. Pretreatment of the specimen was carried out by operation manual. After that, we used UHPLC (1290 Infnity LC, Agilent Technologies) coupled with a quadrupole time-of-fight (AB Sciex TripleTOF 6600) to perform analyses. After uploading the processed data to Metab-oAnalyst (version 4.0, https://www.metaboanalyst.ca) for further analysis, we used principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) on the positive and negative models. Te fnal processing of datasets and analysis of results referencing published literature [32]. Ten, the correlation analysis between gut microbiota and metabolomics was followed as reported [33]. Diferentially abundant gut microbiota and metabolites were log 2 scaled (TMT/iTRAQ) or Z-score scaled (Label-free) and concatenated into one matrix. Ten, correlation coefcient among all the molecules in the matrix was calculated with the Pearson algorithm in R (version 3.5.1, https://www.r-project.org). Pearson correlation coefcient among the diferentially expressed gut microbiota and metabolites was loaded into Cytoscape (version 3.5.1, https://www.cytoscape.org) and the correlation network was calculated.

Statistical Analysis.
Pooled data are presented as the mean ± SEM. Using SPSS Statistics 24 (IBM, Armonk, New York, United States) for one-way ANOVA followed by Tukey's test to fnd a signifcant diference between the tested groups (P < 0.05).

Reduction of HBV Clinical Symptoms by C. nutans.
We observed a dramatically increasing (P < 0.001) in ALT activity and HBsAg levels in serum in the modeling group compared with the control group (see Figures 1(a) and 1(b)). It demonstrated that the HBV model was successfully induced by using HI. To determine whether C. nutans had alleviative efcacy on HBV symptoms, HBsAg, IL-1β, and TNF-α levels in the serum were measured by ELISA, and cccDNA levels in the liver tissues were detected by qPCR, the results show that the C. nutans treated group could be observed a signifcant decline of HBsAg (P < 0.001), IL-1β(P < 0.05), TNF-α(P < 0.01) levels, and HBV cccDNA (P < 0.001) levels compared with the modeling group (see , showing equivalent or better suppressive efects on HBV than Entecavir.

Changes of the Intestinal Microbial Community by C. nutans.
We performed the PCA method, which is based on the relative abundance of genera, to obtain an overview of the microbiota composition between the modeling group mice and the nondisease group, and the results revealed that the group of YDC and ETC, separated from the model group mice (see Figure S1a), showed a signifcant diference in gut microbiota in each group. Te gut microbiota diversity of the mice is shown in Figure S1c-S1f, the results show that the diversity index of Shannon, chao1, observed_species, and PD_whole_tree was higher in the NOR, YDC, and ETC groups than that in the MOD group. Specifc gut microbiota changes in MOD and YDC group mice were accessed across levels of phylum, family, and genus. Te MOD group mice had a signifcant increase of Firmicutes (P < 0.001) and a decrease of Bacteroidetes (P < 0.001) compared with the NOR group mice at the phylum level (see Figure 3(a)). Gut microbiota communities of the mice in the group of YDC and ETC showed a trend of increase of Bacteriodetes (P < 0.001) and decrease of Firmicutes (P < 0.001), compared with those of the MOD group mice (see Evidence-Based Complementary and Alternative Medicine mice had a dramatic decrease of Alistipes(P < 0.05) but a signifcant increase of Lactobacillus(P < 0.05), compared with the NOR group (see Figure 3(h)). Te C. nutans and Entecavir treatment led to a sharp decrease in Lactobacillus(P < 0.001) and led to an increase in Alistipes(P < 0.01) (see Figures 3(i) and 3(j)). Tese results show that the gut microbiota in HBV-model mice was modulated by C. nutans.

Changes of Liver Tissue Metabolites by C. nutans.
Comparing the total ion fow diagram of the QC sample UHPLC-Q-TOF MS with the overlapping spectrograms, the results in Figure S2 show that the response intensity and retention time of each color spectrum peak basically overlap, indicating that the variation caused by the instrument error is small in the whole experiment process. OPLS-DA, based on the relative abundance of positive and negative ion  Figure 3: Efects of C. nutans on the gut microbial community at phylum, family, and genus level. Te heatmap of the microbiota was assessed at the phylum level (a). Data presented are mean ± SEM (n � 6); * * * P < 0.001 for signifcant diference of relative abundance of Firmicutes (b) and Bacteroidetes (c) between various reagents-treated group and model group. Te heatmap of the microbiota was assessed at the family level (d). Data are presented mean ± SEM (n � 6); * P < 0.05, * * P < 0.01, and * * * P < 0.001 for signifcant diference of relative abundance of Lactobacillaceae (e), Bacteroidales_S24-7_group (f ), and Rikenellaceae (g) between various reagents-treated group and model group. Te heatmap of the microbiota was assessed at the genus level (h) and the diference of relative abundance of various bacteria (i, j) between various reagents-treated group and model group.
metabolites, revealed the model group mice showed a distinct microbiota composition from the other groups (see Figures S3a and S3b). Te volcano fgures show the differential metabolites of positive and negative ions between the group of YDC and MOD, and the signifcant diferential metabolites were in pink color (see Figures S2c and S2d). We applied the VIP value >1 and the P < 0.05 analysis to Student's t-test at a univariate level to measure the signifcance of each metabolite and screen 9 common upregulated differential metabolites and 14 common downregulated differential metabolites (see Table 1). Te results showed that C. nutans dramatically upregulate hippuric acid, L-histidine, trehalose, D-threitol, and stachyose and downregulate uridine 5'-diphosphate, cholic acid, trimethylamine N-oxide, CDP-ethanolamine, and phosphorylcholine (P < 0.05). Figure 4(a), Alistipes has a high correlation with the signifcantly diferent metabolites screened out, and it can be seen from the color that Alistipes are positively correlated with upregulated metabolites, such as hippuric acid, Lhistidine, trehalose, D-threitol, and stachyose and showed negativity related with downregulated metabolites, such as uridine 5′-diphosphate, cholic acid, trimethylamine Noxide, CDP-ethanolamine, and phosphorylcholine. In Figure 4(b), Alistipes is the genus with the most signifcant correlation of the signifcantly diferent metabolites screened, showing a signifcant negative correlation with uridine 5′-diphosphate, trimethylamine oxide, CDP ethanolamine and phosphorylcholine (P < 0.05), a signifcant negative correlation with cholic acid (P < 0.01) a signifcant positive correlation with hippuric acid (P < 0.01), a signifcant positive correlation with L-histidine, trehalose, Dthreitol, and stachyose (P < 0.05). In Figure S4, Alistipes is in the most prominent position, suggesting that it is a crucial intestinal microbe. In Figures 4(c) and 4(d), the degree of dispersion shows the distribution of Alistipes and diferential metabolites in the two samples. Te metabolite rho value of Alistipes positive regulation is hippuric acid (Rho � 0.868, P � 1.33e-5, see Figure 4(c)). Te negatively regulated metabolite rho has the largest absolute value for cholic acid (rho � −0.656, P � 0.0058, see Figure 4(d)).

Discussion
C. nutans, as a traditional herb, shows anti-infammatory activity, immunomodulating activity, and antiviral activity [34], but its efect on HBV has rarely been reported. In this paper, we observed its anti-HBV infection efcacy. C. nutans directly interfered with the production of HBsAg, IL-1β, TNF-α in serum, and HBV cccDNA in the liver after analyzing all the experimental data. C. nutans also shows the impact of changing the gut microbial community and liver metabolites in HBV model mice. In addition, from the monitoring results of the body weight and health conditions of mice in all groups, C. nutans is a biosafety and low-cost green medicinal plant that can be used for the long-term use.
C. nutans has noticeable efcacy in alleviating symptoms of HBV infection. Previous studies demonstrated that continued high levels of HBsAg are accompanied by low immune function in chronic hepatitis B (CHB) patients, and the longterm high titer of HBsAg may cause the degree and function of HBV-specifc immune cells to become resistant [35,36]. In this work, C. nutans can signifcantly reduce the serum HBsAg level of model mice, showing the potential to alleviate the failure of immune cell function and carry out immune regulation. In addition, it is reported that elevated levels of infammatory cytokines, such as TNF-α and IL-1β, could be observed in the serum of HBV-infected persons, promoting liver infammation and malignant progression [37][38][39]. As displayed in our results, C. nutans signifcantly reduced the serum levels of TNF-α and IL-1β in model mice and showed an excellent antiinfammatory efect and indicated that the efcacy of C. nutans on HB progression could partly be attributed to its ability to inhibit the secretion of infammatory cytokines. Moreover, it is necessary to eliminate the nuclear cccDNA from the infected liver cells for long-term antiviral therapy [40,41]. In the present study, C. nutans can signifcantly inhibit the level of cccDNA in mouse liver cell nuclear and exhibits excellent antiviral performance and long-term efcacy. However, further research and more comprehensive data are needed to reveal the mechanism that has not been clarifed yet. In short, our results proved that C. nutans exhibits signifcant efcacy in inhibiting HBV infection and HB progression.    Evidence-Based Complementary and Alternative Medicine Investigating the gut microbiota may provide a new perspective into the origination and development of HB. As demonstrated in the past research, the imbalance of the gut microbiota is closely related to the development of CHB disease [8]. After acute or chronic HBV infection in mice, the richness of the Bacteroidetes phyla signifcantly declined. In contrast, the richness of Firmicutes dramatically increased considerably, indicating that the ratio of Firmicutes/ Bacteroidetes (F/B) signifcantly increased dramatically after HBV infection [42]. In our work, the modeling mice had a signifcant increase of Firmicutes and a decrease of Bacteroidetes, further confrming the successful establishment of the HBV mouse model from another aspect. Gut microbiota communities of the mice in C. nutans and Entecavir treated groups showed a trend of increase of Bacteroidetes and a decrease of Firmicutes, indicating that C. nutans may intervene in the progression of HB by changing the ratio of F/B at the phylum level of gut microbiota, as same as Entecavir do.
In addition, C. nutans can signifcantly reduce the relative abundance of Lactobacillaceae at the family level and Lactobacillus at the genus level, shows obvious inhibitory regulation on Lactobacillus-ssp. As reported in the literature that Lactobacillus-ssp has obvious pathogenicity and can cause endocarditis and meningitis [43]. In addition, Lactobacillus paracasei and Lactobacillus rhamnosus isolated from the drainage tube of patients are closely correlated with infections, and Lactobacillus gasseri can be both pathogenic and colonizing bacteria [44]. Although the role of these bacteria in the progression of HB has not been confrmed, it provides a guide for us to further explore the efect of C. nutans on intestinal fora against infectious diseases including HB. Moreover, compared with the modeling group, C. nutans can signifcantly increase the abundance of Rikenellaceae at the family level and Alistipes at the genus level in the intestines of mice. Alistipes belongs to the Bacteroidetes phyla and Rikenellaceae family, indicating that C. nutans may directly act on increasing the quantity of Alistipes to achieve the upregulation of Bacteroidetes and fnally restore the ratio imbalance of F/B.
Metabolomics research provides a new weapon for us to evaluate the body's condition from disease to recovery. As shown in our results, the upregulated diferential metabolites shared by the group of YDC and NOR and/or ETC include hippuric acid and L-histidine. Tese metabolites may be the target metabolites of C. nutans to restore the original level of metabolites in the body. Te unique upregulated metabolites of the YDC group are mainly trehalose, D-threitol, and stachyose. Tese metabolites may have the ability in the transformation of the practical components of the grass. Furthermore, the downregulation of the diferential metabolites, shared by the group of YDC and NOR and/or ETC, including uridine 5′-diphosphate, cholic acid, trimethylamine oxide, CDP ethanolamine, and phosphorylcholine. Te above data confrm that C. nutans exhibits the potential of leading the decrease of oxidative stress in the body and the production of TNF-α in the liver by inhibiting the level of these harmful metabolites [45].
Te correlation analysis between gut microbiota and liver metabolomics helps us screen out the target-regulated bacteria and the critically related metabolites by C. nutans on HB. As displayed in Figure S4 and Figure 4, Alistipes is a crucial intestinal microbe, which is consistent with the results of intestinal fora screening. Te results of Figures 4(c) and 4(d) show that the positive regulation metabolite of Alistipes is hippuric acid while the negatively regulated one is cholic acid, agrees with the results of the moderation trends both in diferent gut microbiota and in diferent liver metabolites. Te published literature revealed that in healthy people, the abundance of Alistipes is signifcantly higher than in HBV-infected patients [46]. When HBV infection progresses to cause a more severe disease stage, it is reduced dramatically, other studies have also shown that CHB patients had a signifcant reduction in Alistipes compared to healthy patients [47], which implies the potential benefts of Alistipes. Hippuric acid is one of the metabolites derived from intestinal microbes, which can be used as an indicator of microbial diversity, and its derivatives have anti-infammatory and anti-HBV activity [48,49]. Cholic acid are considered an interventive targets for relieving liver infammation as well as a potential biomarker for targeted therapy of HBV-infected patients [50][51][52]. All the results imply that the content of Alistipes is increased through the regulation of C. nutans and then directly afects the metabolism levels of hippuric acid and cholic acid, which is the key potential metabolic pathway for the efcacy of C. nutans on HB.

Conclusion
In conclusion, C. nutans is a reliable drug for the prevention and treatment of HB, it exerts a drug efect that alleviative the HBV symptoms in mouse models and has remodeled gut microbiota in mice and modulate the level of liver metabolites to prevent and treat HB. C. nutans shows ability in regulating the crucial bacteria Alistipes, and then restrict the metabolism levels of hippuric acid and cholic acid to play a specifc role in preventing and curing HB.

Data Availability
Te data used to support the fndings of this study are available from the corresponding authors upon request.

Ethical Approval
Te study was approved by the Ethics Committee of Hainan Medical University for Animal Care and Use.

Supplementary Materials
Tis supplementary material included all the research results mentioned in the manuscript. Figure S1: efects of C. nutans for the HBV mouse model on the gut microbial community. Figure S2: positive/negative ion mode TIC overlapping Atlas of samples. Figure S3: efects of C. nutans for the HBV mouse model on the liver tissue metabolites. Figure S4: degree value analysis of very important bacteria. (Supplementary Materials)