Gut microbiota disorder caused by diterpenoids extracted from Euphorbia pekinensis aggravates intestinal mucosal damage

Abstract Gut microbiota disorder will lead to intestinal damage. This study evaluated the influence of total diterpenoids extracted from Euphorbia pekinensis (TDEP) on gut microbiota and intestinal mucosal barrier after long‐term administration, and the correlations between gut microbiota and intestinal mucosal barrier were analysed by Spearman correlation analysis. Mice were randomly divided to control group, TDEP groups (4, 8, 16 mg/kg), TDEP (16 mg/kg) + antibiotic group. Two weeks after intragastric administration, inflammatory factors (TNF‐α, IL‐6, IL‐1β) and LPS in serum, short chain fatty acids (SCFAs) in feces were tested by Enzyme‐linked immunosorbent assay (ELISA) and high‐performance liquid chromatography (HPLC), respectively. The expression of tight junction (TJ) protein in colon was measured by western blotting. Furthermore, the effects of TDEP on gut microbiota community in mice have been investigated by 16SrDNA high‐throughput sequencing. The results showed TDEP significantly increased the levels of inflammatory factors in dose‐dependent manners, and decreased the expression of TJ protein and SCFAs, and the composition of gut microbiota of mice in TDEP group was significantly different from that of control group. When antibiotics were added, the diversity of gut microbiota was significantly reduced, and the colon injury was more serious. Finally, through correlation analysis, we have found nine key bacteria (Barnesiella, Muribaculaceae_unclassified, Alloprevotella, Candidatus_Arthromitus, Enterorhabdus, Alistipes, Bilophila, Mucispirillum, Ruminiclostridium) that may be related to colon injury caused by TDEP. Taken together, the disturbance of gut microbiota caused by TDEP may aggravate the colon injury, and its possible mechanism may be related to the decrease of SCFAs in feces, disrupted the expression of TJ protein in colon and increasing the contents of inflammatory factors.


| INTRODUC TI ON
Euphorbia Pekinensis (EP), the radix of Euphorbia pekinensis Rup, is a well-known Chinese herb which has been used to treat gonorrhea, edema, ascites, migraines, and warts for 1000 years. 1,2 Pharmacological studies show that EP also has anti-tumor and antiangiogenic activities, 3,4 and the active ingredients are some watersoluble high polar substances. However, it is worth noting that EP can cause abdominal pain and diarrhea when used improperly. 5,6 Many studies have shown that the toxic components of EP are concentrated in the low polar parts, especially diterpenoids, which are considered to be the main toxic components. In vitro study, casbane diterpenoids from EP are toxic to many cell lines, such as LO-2, 7 IEC-6 8 and MDCK, 9 they can induce cell apoptosis, cellular morphological change, ROS accumulation, and mitochondrial membrane potential (MMP) disruption. [7][8][9] In our previous study, we demonstrated the toxic effects of casbane diterpenoids in vivo, the possible mechanism is due to the disordered expression of aquaporin in intestinal tract caused by diterpenoids from EP, 5

and inflammation
aggravates the disorder of aquaporin expression. 10 All these studies have proved the toxic effects of diterpenoids from EP; however, the mechanism is still unclear, particularly, the existing research has not paid attention to gut microbiota, which plays an important role in the homeostasis of the gut.
Gut microbiota has been found to be related to many diseases, such as diabetes, 11 cancer, [12][13][14] chronic kidney disease, 15,16 and Parkinson's disease, 17 the gut microbiota regulates the disease process through some key metabolites, thus, homeostasis of gut microbiota is very important to maintain the health of the body. The intestinal mucosal barrier consists of mechanical barrier, chemical barrier, biological barrier, and immune barrier, 18  and epithelial defense. 20 Among these SCFAs, BUT has been the most widely studied, it shows that butyrate induces actin-binding protein synaptopodin (SYNPO) expression in epithelial cell lines and murine colonic enteroids through mechanisms possibly involving histone deacetylase inhibition, which reveals a direct mechanistic link between microbiota-derived butyrate and barrier restoration. 21 All these findings suggest that gut microbiota is essential for the integrity of intestinal mucosal barrier function, and we speculated that the toxic diterpenes from Euphorbia pekinensis may cause severe intestinal mucosal damage by affecting gut microbiota.
In our previous study, we tested the acute toxic of total diterpenoids extracted from Euphorbia pekinensis (TDEP), and most studies on intestinal toxicity of Euphorbia pekinensis focus on acute toxicity; however, in many cases, it needs to be taken for a long time, and the toxicity under this condition is unclear. In this study, 16SrDNA sequencing was used to detect the difference of gut microbiota in mice after TDEP administration for 2 weeks, and seek for different microbiota. Histopathological section of colon and the TJ protein expression was tested to confirm the damage of intestinal mucosal barrier. The content of SCFAs in intestinal feces was also determined, and we further used antibiotic interference to verify the toxic effects of TDEP and the protective effects of SCFAs. Our findings reveal that gut microbiota disorder caused by TDEP aggravates intestinal mucosal damage.

| Preparation of TDEP
TDEP were extracted and isolated from the radix of EP according to previous studies. 5 6 known diterpenoids accounting for 85.26% of TDEP. TDEP were dissolved in methanol and detected using HPLC.

| Measurement of serum levels of inflammatory cytokines and lipopolysaccharide (LPS)
At the end of the treatment, all mice were sacrificed by cardiac puncture under 10% chloral hydrate (0.7 ml/100 g, i.p.). Blood

| Histopathological examination
The colon samples (n = 6) were fixed in 10% phosphate-buffered formalin, dehydrated and then embedded with paraffin. Subsequently, the tissues were cut into 5μm sections and stained with hematoxylineosin (HE). Representative micrographs of the colon sections were obtained using a 400× objective under a light microscope.

| Determination of SCFAs using high performance liquid chromatography (HPLC)
Briefly, 300 mg feces that was added to 1 ml ultra-pure water and  Bioanalyzer (Agilent) and Illumina (KapaBiosciences) library quantification kits. The gut microbiota profile was determined using a MiSeq high-throughput sequencing platform (NovaSeqPE250).

| Statistical analysis
SPSS version 16.0 for Windows (SPSS) was used for statistical analysis. Numerical data were expressed as mean ± SD. The significance of differences was examined using one-way analysis of variance (ANOVA) procedure followed by the Dunnett's test. The correlations between microbiota and host parameters were analyzed by Spearman's correlation. Results with p < .05 were considered statistically significant.

| Histological results
Representative HE staining of colon tissues is shown in Figure 1A. The pathological morphology of control group was normal, no inflammatory response and damage. In the high dose group of TDEP, significant The results were normalized with β-actin protein level, and all TJ proteins level of the control was taken as 100%. Data are represented as the mean ± SD. *p < .05, **p < .01 versus control group, respectively. n = 6 in each group and each assay was repeated three times mucosal damage was observed, the villi were irregular with local epithelial shedding, inflammatory infiltration of large areas of mononuclear leukocytes in the mucosa and submucosa, and there was edema between mucosal and muscular layers in the colon. In the antibiotic treatment group, mucosal damage was more serious. Inflammatory infiltration was also observed in the medium dose group, some of the epithelial cells fell off, and the edema between the colonic mucosa and muscular layer was alleviated, and the damage of low dose group was not obvious compared with control group.

| Effects of TDEP on expression of inflammatory factors and TJ protein levels
The damage of mucosal barrier function may lead to inflammatory reaction, in this study, the expression of IL-1β, IL-6, and TNFα in serum was tested by enzyme linked immunosorbent assay (ELISA). The result shows that TDEP lead to the increase of inflammatory factors in blood in a dose-dependent manner(p < .05), and the expression of inflammatory factors was even higher than TDEP high dose group in antibiotic treatment group, content of LPS in serum was also detected, the result was consistent with the inflammatory factors expression (p < .05).
Occludin, claudin-1, and ZO-1 are important tight junction proteins, they are critical for the maintenance of intestinal mucosal barrier function. The expressions of these three TJ proteins in colon were detected, as shown in Figure 1C, after TDEP administration, the expression of TJ protein in colon of mice was significantly decreased, and the expression of TJ protein in the colon of antibiotic treatment group was lower than that of high dose group (p < .05).

| The contents of SCFAs in feces
The contents of SCFAs in mice feces at different doses of TDEP were tested. As shown in Figure 2, acetic acid, propionic acid, and butyric acid are the main SCFAs in feces, accounting for about 80% of SCFAs. In low dose of TDEP(4 mg/kg), there was significant difference in the contents of acetic acid, i-butyric acid, n-butyric acid, and hexanoic acid (p < .05). In medium dose group, all the SCFAs were decreased (p < .05). While in high dose group, all SCFAs were significantly decreased (p < .05). Compared with high dose group, acetic acid, i-butyric acid, and hexanoic acid were significantly decreased in antibiotic group (p < .05), and n-butyric acid was not detected in high-dose and antibiotic groups.

| 16SrDNA sequencing
These sequence data have been submitted to the Sequence Read Archive(SRA) databases under accession number SUB8556324.

| Alpha and Beta diversity analysis
Alpha diversity analysis is used to evaluate the species diversity of different treated groups, which includes the Chao1, Observed species, Goods_coverage, Shannon, and Simpson indexes. In this research, Chao 1 index and Goods_coverage index results showed that species value was significantly different when antibiotics were used. However, there was no significant difference between control group and TDEP group. Community diversity was estimated using The results showed that TDEP group changed the gut microbiota significantly from control group, and antibiotic treatment group showed an obvious separation of other two groups. These indicated that there were significant differences in gut microflora among the three groups ( Figure 3B).

| The microbial community structures at the phylum and genus levels
From the diversity results, we can see that the species composition was quite different between three groups. We further selected the highest abundance from the phylum and genus level to analyze the species differences among three groups. As illustrated by Figure 4A The data showed that there was significant difference between the antibiotic group and the other two groups, but there was no significant difference between TDEP group and control group. (B) Weighted and unweighted PCoA analysis. The PCoA analysis showed a clear separation of the TDEP group from the control group and the antibiotic treatment group species cannot even be detected. In the cluster analysis of the bacterial phyla and genera, it was found that the distributions of these phyla and genera in the TDEP group were closer to that in the control group than in antibiotic group.

| LEfSe analysis in TDEP and antibiotictreated group
To define which bacterium might be responsible for colon injury induced by TDEP, linear discriminant analysis (LDA) effect size (LEfSe) was used to analyze the differences among the three groups from the phylum level to the genus level, and the magnitudes of effects of the different species biomarkers were assessed by LDA ( Figure S1).
The results showed that gut microbiota differed significantly among three group, about 59 biomarkers were found ( Figure 5). We removed the bacteria with relative abundance less than 0.1% for further analysis, and 35 bacterial genera were selected. The correlations between the 35 bacterial genera and biochemical parameters were analyzed by Spearman's correlation analysis, eventually, 9 bacteria genera with significant correlation with some biochemical parameters are shown in Table 1. Notably, in antibiotic-treated group, some opportunistic pathogens such as Klebsiella were detected, and this may be one of the reasons why it is more toxic.

| Correlation between the abundances of different bacterial genera and biochemical parameters
As shown in Table 1, ten bacterial genera with significant changes in abundance after TDEP administration were selected, and the relation-

IL
.049 .004 .072 .023 .058 Significant correlations (p < .05) are in bold. harmful substances were more easily to enter the blood. The conjecture was verified in further experiments, as shown in Figure 1B, the level of LPS in serum of high dose group was nearly three times that of control group. As we all known, LPS is product of Gramnegative bacteria, which induce inflammatory reaction through Toll-like receptor, leading to intestinal dysfunction and even other organ damage. 24 We also found that the expression of inflammatory factors also showed a dose-dependent relationship, all these results were consistent with the histopathological damage. However, the mechanism of TDEP induced colon injury remained unclear.
Gut microbiota has been shown to be associated with tight junction proteins expression and is essential for maintaining intestinal physicochemical barrier. 25 To investigate whether gut microbiota played a role in TDEP induced intestinal toxicity, antibiotics were added to interfere with the gut microbiota when TEDP was given to mice. It was worth noting that the damage of colon tissue was more obvious when antibiotics were added, the level of inflammatory factors in serum increased, and the expression of tight junction proteins was lower. These indicated that when the abundance of some bacteria was inhibited by antibiotics, the protective effect of bacteria on intestinal mucosa was also reduced, which eventually lead to the aggravation of colon injury. To find out the potential microbiota which associated with biochemical parameters closely related to colon injury, 16SrDNA high-throughput sequencing technology was used to study the changes of gut microbiota after TDEP intervention. The results showed there were gut microbiota disorders in TDEP group.
Although there was no significant change in alpha diversity of gut microbiota after intragastric administration of TDEP, there were significant differences in the gut microbiota abundance between TDEP groups and control group. PCoA analysis showed that the three groups could be significantly separated. In the species analysis, we first analyzed the difference of microbiota at phy- Gordonibacter in smoking patients with Crohn's disease was reduced compared with nonsmoking patients. 33 In another study, GFP-Cr could significantly increase the relative abundance of Enterorhabdus in diabetes mellitus mice. 34 These findings suggested that Enterorhabdus may play a positive role in maintaining intestinal barrier function.
Candidatus Arthromitus 35 was proved closely related to the intestinal mucosal immunity of the host, which promoted immune maturation and enhances host resistance, and it was difficult to recovery after large doses of antibiotics treatment. All these bacteria genera were not detected in antibiotic group, which may be one of the reasons for the strong toxicity of the antibiotic group.
On the other hand, the abundances of Mucispirillum, Bilophila, Ruminiclostridium increased after TEDP adminstration. Zhang, et al. 36 found that Shen-Ling-Bai-Zhu-San could improve functional dyspepsia by reducing functional dyspepsia biomarkers including Prevotella, Mucispirillum, and Akkermansia. Bilophila wadsworthia, which belong to the Bilophila genera, had been proved to promote higher inflammation, intestinal barrier dysfunction and bile acid dysmetabolism, leading to higher glucose dysmetabolism and hepatic steatosis, 37 and it had also been confirmed to be associated with colorectal cancer. 38 The increase of these pathogens is another cause of colon injury caused by TDEP. What's more, Klebsiella and Escherichia-Shigella were found in antibiotic group, both of them were pathogenic bacteria, but they were not detected in other two groups. We speculated that antibiotics inhibit the original gut microbiota, which made these pathogens colonize more easily and caused serious damage.
Moreover significant different functional profiles between different groups were predicted by PICRUSt. As the results showed, decrease of glyoxylate cycle and TCA cycle were observed in TDEP group, the glyoxylate cycle was a variation of the TCA cycle, and they were the hub for energy metabolism, the decrease of them meant that the energy metabolism of mice was reduced. Biosynthesis of ubiquinol 7−10 was also significantly reduced, ubiquinol also called coenzyme Q, was a well-known antioxidant molecule, the reduction of them was not conducive to the development of antioxidant defenses in colon. There were also some metadata pathways changed after TDEP administration. However, whether they are the cause of TDEP induced colon injury need further study.

| CON CLUS IONS
The experiment confirmed for the first time that colon injury induced by TDEP is associated with disturbance of gut microbiota. Through the determination of inflammatory factors in serum, tight junction proteins in colon and short chain fatty acids in feces, the damage of TDEP to colon was confirmed. The colon injury was more obvious when antibiotics were added, which suggested that some gut bacteria might play an active role in TDEP-induced colon injury. Through the correlation analysis, 7 bacteria that are beneficial to colonic function were identified (Barnesiella, Muribaculaceae_unclassified, Alloprevotella, Candidatus_Arthromitus, Enterorhabdus, Alistipes), and 3 bacteria that were harmful to colonic function were found (Bilophila, Mucispirillum, Ruminiclostridium). This study reminded us that attention should be paid to the changes of gut microbiota when using traditional Chinese medicine for a long time to avoid adverse effects. The regulation of these bacteria can improve intestinal diseases such as inflammatory bowel disease.

ACK N OWLED G M ENTS
This work was financially supported by the National Natural Science

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are opened available in SRA (SUB8556324).