Effects of Xiao-Ban-Xia-Tang Formula on Cisplatin and 1-phenylbiguanide Hydrochloride Induced Emesis and Gut Microbiota in the Pica Model of Rats

Background: In this study, the effects of Xiao-Ban-Xia-Tang (XBXT) formula on cisplatin and 1-phenylbiguanide hydrochloride (1-PBG) induced acute and delayed emesis and gut microbiota were studied in the pica model of rats, compared with ondansetron. Methods: Two rat models of cisplatin and 1-PBG induced pica were established, the amount of kaolin intake was observed, and the effects of XBXT and ondansetron on the gut microbiota were further studied by 16S rDNA gene analysis. Results: The results showed that the total intake of kaolin of the rats injected with cisplatin and 1-PBG was signicantly increased, and treatment of XBXT and ondansetron could signicantly ameliorate the acute and delayed pica induced by cisplatin and 1-PBG. The 16S rDNA gene analysis has shown that the alpha diversity of the gut microbiota of the cisplatin and 1-PBG treated rats was signicantly decreased compared with the control group, and ondansetron could further decrease the alpha diversity of the gut microbiota of the rats treated with cisplatin. Ondansetron signicantly decreased the relative abundance of Firmicutes and increased the abundance of Bacteroidetes on phylum level in the cisplatin and 1-PBG treated rats, while XBXT only decreased Firmicutes in the cisplatin treated rats. Conclusions: XBXT was as effective as ondansetron in the treatment of acute and delayed pica induced by cisplatin and 1-PBG in rats. Ondansetron was more likely to cause gut microbiota dysbiosis than XBXT. Our study provided new avenues for the roles and mechanisms of XBXT on the prevention and treatment of CINV.


Background
Nausea and vomiting is one of the most common and painful symptoms in humans. Gastrointestinal disease, endocrine or metabolic diseases, surgery, neurological or psychiatric factors, pregnancy and some medicines can cause nausea and vomiting. Chemotherapy-induced nausea and vomiting (CINV) is one of the severe adverse effects of chemotherapy, which can affect the quality of life and compliance of cancer patients. CINV can occur in the acute phase (0-24 h) and in the delayed phase (25-120 h) after chemotherapy, and many neuroreceptors such as dopamine, serotonin, and neurokinin-1 (NK1), have been found to be involved in the pathological process of chemotherapeutic vomiting [1,2]. The main medicines used for the treatment of CINV include 5-Hydroxytryptamine (5-HT3) receptor antagonists, NK1 receptor antagonists, corticosteroids, dopamine antagonists, benzodiazepines, cannabinoids and olanzapine [3]. These medicines usually act on a single target, and are often needed to be combined with other medicines for the treatment of CINV except for expensive costs and side effects, such as constipation, headache and hiccups, et al [4]. In addition, it has been reported that delayed CINV has a higher incidence and is less responsive to treatment compared with acute CINV [1,2]. So it is still urgent to nd potential medicines for the prevention and treatment of acute and delayed CINV.
Traditional medicine converts traditional theories and experiences into medical knowledge and skills to prevent and treat diseases, and has been recommended by the World Health Organization as complementary and alternative medicine for its effect, safety, availability and low cost [5]. Xiao-Ban-Xia-Tang (XBXT) formula, a traditional Chinese medicine recipe, has been used in China for more than 2000 years. XBXT includes only two herbal medicines, ginger (Zingiber o cinale Roscoe) and pinellia (Pinellia ternata ( Thunb.) Breit.), and is usually used for the treatment of vomiting [6]. Ginger and its extracts have been found to have many biological and pharmaceutical activities, such as anti-in ammatory, anticancer, antiobesity, antinausea and antiemetic activities [7]. Our previous study has shown that [6]-Gingerol, a major component of ginger, could signi cantly ameliorate the cisplatin-induced pica in rats [8]. Pinellia species also have many biological activities including antiin ammatory, anticancer, insecticidal, antitussive, anticonvulsant and antiemetic activities [9]. The main active components of Pinellia include alkaloids, lectins, fatty acids, cerebrosides, phenylpropanoids, sterols and avonoids [9. 10]. Qian Q et al reported that XBXT could ameliorate cisplatin-induced emesis in minks by inhibiting the increase of Neurokinin-1 receptor (NK1-R) [6]. Qian Q's another study reported that XBXT could improve cisplatin-induced pica in rats by inhibiting the increase of central or peripheral obestatin and blood Cholecystokinin (CCK) and Calcitonin gene-related peptide (CGRP) [11]. Since traditional Chinese medicine recipe usually have many ingredients and multi-targets, the mechanism of XBXT remain not clear yet.
In our previous study, the effects of ondansetron and [6]-gingerol on pica and gut microbiota in rats injected with cisplatin were evaluated [8]. In the present study, two nausea and vomiting models of rats have been established: one was induced by cisplatin, and the other was induced by 1-phenylbiguanide hydrochloride (1-PBG), which is an agonist of 5HT3 [12]. The pica behavior, consumption of kaolin (china clay), was used as an indicator of nausea and vomiting. Both cisplatin and 1-PBG caused acute and delayed behavior of pica. Ondansetron and XBXT were used to treat the two models of rats, and the effects of Ondansetron and XBXT on the gut microbiota of the cisplatin and 1-PBG treated rats were further checked by 16S rDNA gene analysis, providing new sights for the prevention and treatment of CINV.

Preparation of kaolin pellets
The preparation of kaolin pellets was described in our previous work [8].

Preparation of XBXT
Ginger (Zingiber o cinale Roscoe, produced in Laiwu, Shandong Province, China, voucher specimen no. J7201) and pinellia (Pinellia ternata ( Thunb.) Breit., produced in Xihe County, Gansu Province China, voucher specimen no.J7654) were identi ed by Professor Jizhu Liu from School of Chinese Materia Medica, Guangdong Pharmaceutical University. Voucher specimens were deposited at the Herbarium of Traditional Chinese Medicine of Guangdong Pharmaceutical University.
Ginger 200g and pinellia 400g were selected and weighed precisely. The herbs were soaked in 4800 mL distilled water overnight, and back-owed for 1.5 h. The extract was ltered, and the residual medicine was soaked in distilled water following the same procedure once more. The pool of the extracts from two soaking and ltering was lyophilized to form a dried powder. -shogaol, 0.3 g dried powder of XBXT was dissolved in 1 mL petroleum ether, and ltered through a 0.45 μm lter (Jinteng Experimental Equipment Co., Ltd., Tianjin, China). The HPLC analysis was performed using the HPLC system (U3000, Thermo Fisher Scienti c) with acetonitrile as mobile phase A and 0.1% formic acid solution as mobile phase B at 280 nm. The ow rate was 0.5 mL/min and the column temperature was 30 ˚C. For the check of ephedrine, 0.3 g XBXT powder was dissolved in 1 mL 75% ethanol (containing 1% hydrochloric acid). The mobile phase was methanol: 0.08% triethylamine aqueous solution (pH=5.8). The ow rate was 1.0 mL/minute and the detection wavelength was 210 nm. For the check of succinic acid, 0.3 g XBXT powder was dissolved in 1 mL petroleum ether. The mobile phase was methanol: 0.02 mol/L KH 2 PO 4 buffer (pH=2.5). The ow rate was 1.0 mL/minute and the detection wavelength was 214 nm. Petroleum ether, acetonitrile, formic acid, ethanol, hydrochloric acid, methanol, triethylamine and KH 2 PO 4 were purchased from Zhiyuan Chemical Reagent Co., Ltd., Tianjin, China. The results of HPLC analysis were shown in Figure S1-3.

Animal treatment
All the animal experiments were approved by the Committee on Laboratory Animal Care and Use of Guangdong Pharmaceutical University (Guangzhou, China), in accordance with the National Institutes of Health guide for the care and use of laboratory animals. The Wistar rats (180-220g) (provided by Peng Yue experimental animal Co., Ltd., Jinan, China) were housed in a temperature-controlled room (25 ˚C), and the relative humidity was 40%-60%.
The kaolin pellets were introduced into the rats 3 days prior to drug administration. Most of the rats did not take kaolin anymore at the third day, and the rats that were still interested in kaolin were excluded. The rest rats were randomly divided into seven groups and each group had 6 rats. Group 1 was the normal control group (Control); group 2 was the rats treated with cisplatin (C-Model); group 3 was the rats treated with cisplatin and ondansetron (C-Ond); group 4 was the rats treated with cisplatin and the XBXT (C-XBXT); group 5 was the rats treated with 1-PBG (P-Model); group 6 was the rats treated with 1-PBG and ondansetron (P-Ond); group 7 was the rats treated with 1-PBG and XBXT (P-XBXT).
On the day of drug administration, the C-Ond group and the P-Ond group were given ondansetron (1.3 mg/kg (body weight), Qilu Pharmaceutical Company, China) by gavage respectively. The C-XBXT group and the P-XBXT group were given XBXT (0.16 g/mL) by gavage respectively. The control group, the C-Model group and the P-Model group were given pure water by gavage. After 1 h, the C-Model group, C-Ond group and the C-XBXT group were injected intraperitoneally (i.p.) with cisplatin (Qilu Pharmaceutical Company, China) at the concentration of 6 mg/kg (body weight); the P-Model group, P-Ond group and the P-XBXT group were injected i.p. with 1-PBG (Sigma) at the concentration of 25 mg/kg (body weight), and the control group was injected i.p. with saline of equal volume. The C-Ond group and the P-Ond group were continued to be given ondansetron (1.3 mg/kg body weight) by gavage and the C-XBXT group and the P-XBXT group were continued to be given XBXT (0.16 g/mL) by gavage, twice each day for three days. The general conditions, the body weight and kaolin intake of rats were measured every 24 h, and recorded until 72 h after the establishment of the models.

16S rDNA gene analysis
After the establishment of the models for 72 h, fecal samples were collected. Fecal bacterial DNA extraction, 16S rDNA gene PCR ampli cation, sequencing and analysis were carried out by Gene Denovo Biotechnology Company (Guangzhou, China). The experimental procedures were performed as previously described [8,13].

Statistical analysis
Statistical differences were determined by using SPSS 23.0 software (SPSS Inc., Chicago, IL, USA). Data were expressed as mean ± S.E. Two-way ANOVA was performed when more than two groups were compared, P value less than 0.05 was considered to be signi cant.

Results
Effect of XBXT on cisplatin and 1-PBG induced kaolin intake All the rats took kaolin at the rst day of kaolin release, and the amount of kaolin intake was gradually decreased at the following 2 days. At the third day, almost all rats did not take kaolin anymore. After injected i.p. of cisplatin, the total intake of kaolin in the C-model group was signi cantly higher than the control group (P < 0.001). The amount of kaolin intake of the C-Model group reached the peak at 24 h, and the second higher amount appeared at 72 h ( Fig. 1A and Table 1). After injected i.p. of 1-PBG, the total intake of kaolin in the P-model group was signi cantly higher than the control group at 24 h (P < 0.001), and the amount reached the highest at 72 h ( Fig. 1A and Table 1). Both ondansetron and XBXT could signi cantly decrease the kaolin intake of the rats treated with cisplatin at 24 h and 72 h, and for rats treated with 1-PBG, both ondansetron and XBXT could also obviously decrease the kaolin intake at 24 h and 72 h ( Fig. 1A and Table 1). Figure 1B demonstrated the total amount of kaolin intake from 0 h (cisplatin and 1-PBG were injected i.p. at 0 h) to 72 h of the 7 groups of rats, and the amount of kaolin intake was signi cantly increased in the C-Model group and the P-Model group (P < 0.01, Fig. 1B). Ondansetron could signi cantly inhibit the kaolin intake of the cisplatin and 1-PBG treated rats (P < 0.05, Fig.1B), and XBXT could inhibit the kaolin intake of the 1-PBG treated rats (P < 0.01, Fig. 1B).
After cisplatin and 1-PBG injection, the body weight of rats in the C-Model group and the C-XBXT group was signi cantly decreased compared with the control group (P < 0.001), and there was no signi cant difference of the body weight between other groups (Fig. 1C, Table 2).
Overview of the 16S rDNA gene analysis The effects of ondansetron and XBXT on the gut microbiota of cisplatin and 1-PBG treated rats were investigated by 16S rDNA gene sequencing. The total tags, operational taxonomic units (OTUs), the Ace index, Chao index, Shannon index and Simpson index were shown in Table 3. The Shannon rarefaction curves for each group have reached the saturation plateau, indicating that the samples had enough sequence coverage (Fig. 2A). The Venn diagram showed that there were 527 common OTUs in the 7 groups of rats (Fig. 2B), there were 597 common OTUs in the control group and the groups treated with cisplatin (Fig. 2C), and there were 637 common OTUs in the control group and the groups treated with 1-PBG (Fig. 2D).
The effects of XBXDT on the gut microbiome of cisplatin and 1-PBG treated rats Figure 3A demonstrated the Chao index of the 7 groups of rats. The alpha diversity of the gut microbiota of the C-Model and P-Model group was signi cantly decreased compared with the control group (P < 0.01 and P < 0.001 respectively), and ondansetron could further decrease the alpha diversity of the gut microbiota of the rats treated with cisplatin (P < 0.01, Fig. 3A). Figure 3B showed the Principal co-ordinates analysis (PCoA) of the 7 groups of rats. Figure 4A-B demonstrated the relative abundance of gut microbiota on phylum level, and Bacteroidetes, Firmicutes, Verrucomicrobia and Proteobacteria had high abundance in all the samples. The abundance of Firmicutes had no obvious change in the C-Model group and the P-Model group compared with the control group, but was signi cantly decreased in the C-Ond, C-XBXT and P-Ond groups compared with the control group (P<0.001, Fig. 4A-B). The abundance of Bacteroidetes of the C-Ond group and the P-Ond group was signi cantly increased compared with the control group (P<0.001, Fig. 4A-B). The abundance of Verrucomicrobia of the C-Model group and the C-XBXT group was signi cantly increased compared with the control group (P<0.05 for the C-Model group, P<0.001 for the C-XBXT group, Fig. 4A-B). The relative abundance of gut microbiota on class level was demonstrated in Figure S4. The abundance of Clostridia was signi cantly decreased in the C-Model, C-Ond, C-XBXT, P-Model, P-Ond and P-XBXT groups compared with the control group (Fig. S4). The abundance of Bacteroidia of the C-Ond group and the P-Ond group was signi cantly increased compared with the control group (Fig. S4). Figure 5A-5B demonstrated the LEFse analysis of the cisplatin treated groups. The speci c and predominant bacteria of the control group included Clostridlales_vadinBB60_group, Lachnospiraceae_NK4A16_group, Angelakisella, Ruminiclostridium_6, Ruminococcaceae_UCG_014, Ruminococcus_2, Candidatus_Saccharimonas, Saccharimonadaceae, Saccharimonadales, Saccharimonadla, Mollicutes_RF39 and Mollicutes, and the speci c bacteria of the C-Model group was Clostridlales_bacterium_CIEAF_020 (Fig. 5A-5B). Figure 5C-5D showed the speci c and predominant bacteria of the 1-PBG treated rats. Clostridiales and Clostridia were speci c for the control group; Candidatus_Saccharimonas, Saccharimonadaceae, Saccharimonadales and Saccharimonadia were speci c for the P-Model group; Prevotella_9, Fusicatenibacter, Phascolarctobacterium, Acidaminococcaceae, Selenomonadales, Negativicutes, Parasutterella, Burkholderiaceae, Betaproteobacteriales and Gammaproteobacteria were predominant in the P-Ond group; Lactobacillus_reuteri, Lactobacillus, Lactobacillaceae, Lactobacillales and Bacilli were predominant in the P-XBXT group (Fig. 5C-5D).

Discussion
In the present study, rats were treated with cisplatin and 1-PBG, and the consumption of kaolin was checked as an indicator of nausea and vomiting. Both cisplatin and 1-PBG could induce acute and delayed emesis. The amount of kaolin intake of the C-Model group reached the peak at 24 h, while the amount of kaolin intake of the P-Model group reached the peak at 72 h. It seemed that cisplatin mainly caused acute emesis, and 1-PBG mainly caused delayed emesis in rats. 5-HT3 receptor antagonists and dexamethasone are the standard anti-emetic regimen for patients receiving highly emetogenic forms of chemotherapy (HEC) such as cisplatin-based regimens; however, delayed emesis remains a problem, since only about half of the cancer patients can get complete control of emesis after chemotherapy [14]. 1-PBG is a 5HT3 agonist, and rats injected i.p. with 1-PBG were induced acute and delayed CINV, moreover, 1-PBG induced CINV mainly presented as delayed CINV. Therefore, 1-PBG provided a good animal model, which could be used for the study of delayed CINV in future.
Many neurotransmitters pathways are involved in the pathological process of CINV, and there is no single neurotransmitter responsible for all forms of CINV, therefore, the prevention and treatment of CINV need the combination of several agents [15]. Although the standard anti-emetic regimen of 5-HT3 receptor antagonists and dexamethasone has been used for many years in patients receiving HEC [14], more and more medicine are being developed for the prevention and treatment of CINV. In the present study, the effects of XBXT, a famous traditional Chinese medicine formula for vomiting, on cisplatin and 1-PBG induced acute and delayed emesis and the gut microbiota were studied, compared with ondansetron, the typical medicine for CINV. Our results showed that both ondansetron and XBXT could signi cantly inhibit the kaolin intake induced by cisplatin and 1-PBG at 24 h and 72 h. Moreover, for the 1-PBG rat model, the amount of kaolin intake of the XBXT group was even lower than the ondansetron group (Fig. 1A-B and Table 1), indicating that XBXT might be more effective in the prevention and treatment of 5HT3 involved CINV.
The gut microbiota has got great interest in recent years since it plays important roles in maintaining health and the development of many diseases [16,17]. Due to the complex composition and complicated mechanism of actions, the potential clinical application of herbal medicine has not been widely recognized yet [18]. With the advancements in genome sequencing technologies and bioinformatics, gut microbiota has become a new avenue to study the traditional medicine [19]. In the present study, the effects of XBXT on the gut microbiota of cisplatin and 1-PBG treated rats were further checked by 16S rDNA gene sequencing analysis, compared with ondansetron. The results of 16S rDNA gene analysis showed that the alpha diversity of the gut microbiota of the C-Model and P-Model group was signi cantly decreased compared with the control group, and ondansetron further decreased the alpha diversity of the gut microbiota of the rats treated with cisplatin. Feng Y et al. reported that Bi dobacterium and Lactobacillus were associated with alpha diversity, beta diversity, and the robustness of the gut microbiota, and people harboring Bi dobacterium present but no Lactobacillus showed higher alpha diversity and were more robust than those only carrying Lactobacillus [20]. Increased microbiome diversity could promote the stability of the microbiome, thereby bene cial to the host for resistance against extreme stress and perturbations [21]. In the present study, both cisplatin and 1-PBG could decrease the alpha diversity of the gut microbiota, while ondansetron could further signi cantly decrease the alpha diversity of the gut microbiota of the rats treated with cisplatin. XBXT did not further decrease the alpha diversity of the gut microbiota of the cisplatin treated rats, therefore, ondansetron might have more severe side effects than XBXT.
In the present study, the abundance of Firmicutes was signi cantly decreased in the C-Ond, C-XBXT and P-Ond groups compared with the control group, while the abundance of Bacteroidetes was signi cantly increased in the C-Ond group and the P-Ond group compared with the control group. Ondansetron signi cantly decreased the relative abundance of Firmicutes and increased the abundance of Bacteroidetes, which was consistent with the result of our previous study [8], while XBXT only decreased Firmicutes in the cisplatin treated rats. Zeng Y et al reported that the abundance of Firmicutes was decreased, and Bacteroidetes was increased in patients with chronic hepatitis, liver cirrhosis and hepatocellular carcinoma compared with the healthy controls [22]. Kim JW et al reported that the Firmicutes/Bacteroidetes ratio was consistently reduced in systemic lupus erythematosus (SLE) patients [23]. On the contrary, Qiu D et al reported that the bacterial communities of the glucocorticoidinduced obesity group of people were increased in Firmicutes and decreased in Bacteroidetes compared with the healthy controls [24]. Zhao L et al reported that the ratio of Firmicutes/Bacteroidetes was higher in type 2 diabetes (T2D) patients than the healthy controls [25]. Therefore, considering the high diversity of gut microbiota, the unexpected microbiota-host relationships and the underlying molecular mechanisms remain to be explored, and food and drugs can profoundly affect the gut bacterial community and change the activities of the gut microbiota [26]. The question "does an altered gut microbiota contribute to disease, or does it merely re ect a disease status?" remains not solved yet [27]. From the results of the present study, it seemed that ondansetron was more likely to cause gut microbiota dysbiosis than XBXT. An X et al reported that the interactions between the gut microbiota and herbal medicines occur mainly by two pathways: one is that the gut microbiota disintegrates the herbal medicines into metabolites and signaling molecules, which induce the physiological changes of the body; the other is that herbal medicines regulate the composition and relative abundance of the gut microbiota, and induce physiological changes [18]. The effects of XBXT on the gut microbiota and the underlying mechanisms still need further study.

Conclusion
In summary, ondansetron and XBXT could signi cantly ameliorate the acute and delayed pica induced by cisplatin and 1-PBG in rats. The alpha diversity of the gut microbiota of the C-Model and P-Model group was signi cantly decreased compared with the control group, and ondansetron could further decrease the alpha diversity of the gut microbiota of the rats treated with cisplatin.
Ondansetron signi cantly decreased the relative abundance of Firmicutes and increased the abundance of Bacteroidetes in the cisplatin and 1-PBG treated rats, while XBXT only decreased Firmicutes in the cisplatin treated rats. Ondansetron was more likely to cause gut microbiota dysbiosis than XBXT. XBXT could be used for the prevention and treatment of CINV, open new avenues to the mechanisms of herbal medicines associated with the gut microbiota.

Consent for publication
All authors agree with the submission.

Availability of data and material
The data of 16S rDNA gene analysis have been uploaded to gshare (https:// gshare.com/), https:// gshare.com/articles/Effects_of_Xiao-Ban-Xia-Tang_formula_on_cisplatin_and_1-phenylbiguanide_hydrochloride_induced_acute_and_delayed_emesis_and_gut_microbiota_in_the_pica_model_of_rats/11316230 Anyone who is interested in the study of gut microbiota, chemotherapy-induced nausea and vomiting, and related elds can download the data for free.

Competing interests
The author reports no con icts of interest in this work.      The alpha diversity and the Unifrac analyses of the gut microbiota in the 7 groups. A: the alpha diversity of the 7 groups; B: the Principal co-ordinates analysis (PCoA) of the 7 groups of rats.